Release Notes

Jul 12, 2017

Cluster (2.6.3)

  • Upgrade Notes
    • Addressed upgrade issue related to internal test automation.

  • Cluster changes
    • Reverted update to the cpanm version used by the Perl stager because it was failing to build XS libraries.

    • Changed aufs log file path to ext4 log file path on logforwarder state restore.

    • Fixed security exposure caused by tracing internal generic data structures in job_manager code.

    • Fixed regression bug so that provider state is not exported to HTTP clients.

    • Implemented logic to truncate local job logs after 10MB is used. Note that the log file offset will continue to increase, but the file size on disk will not exceed 10MB.

    • Added rel/2.6.x documentation to the maintenance release build.

    • Removed hardcoded /data path for NFS mounts.

    • Added delay compress so Splunk finds the log files when IPsec is enabled.

    • Added logic to reclaim virtual network IP address on apc job stop.

    • Added additional monitoring logic to detect if IPsec tunnels are operational.

  • Chef changes
    • Added logic to exit after the first Chef error and not continue processing the next node.

Apcera Setup (2.2.4)

  • No changes.

APC (2.6.3)

  • Added ANSI terminal emulation support on Windows for a better experience, especially when connected to a Linux container.

Web Console (8.0.9)

  • No changes.

Orchestrator (1.0.0)

  • No changes.

Guide (2.6.8)

  • Updated these release notes.

  • Updated the DNS configuration documentation.

  • Updated the Riak configuration documentation.

  • Updated EE installation instructions for Azure.

  • Added documentation versioning with selection menu.

  • Changed the documentation site color scheme (green to blue).

  • Corrected Splunk confiugration description.

  • Updated RAM requirements for Postgres and MySQL docker images.

May 11, 2017

Cluster (2.6.2)

  • Upgrade Notes
    • No changes.

  • Cluster changes
    • Fixed bug related to upgrade where bind from 2.6 job to 2.4.2 network-service fails.

    • Fixed bug where Service Gateways were too strict with the URI.

    • Added start_failure_penalty options for IM taint job schedulig.

    • Updated the package version of the Intel driver.

    • Fixed bug with HTTP router auto-announce functionality.

  • Chef changes
    • Added logic to allow for the forced rotation of HTTP router access logs.

    • Added logic to unhold package libuuid1.

    • Added diagnostics script to cluster deployments for troubleshooting.

    • Added Chef script to install on Gluster servers.

    • Adjusted log file retention to address override bug.

Apcera Setup (2.2.4)

  • No changes.

APC (2.6.2)

  • No changes.

Web Console (8.0.9)

  • No changes.

Orchestrator (1.0.0)

  • No changes.

Guide (2.6.7)

  • Updated Apcera CE installation on GCE docs.

  • Updated the required ports documentation.

  • Added Swagger spec download icon to the TOC, API overview page, and endpoints page.

  • Updated Apcera EE installation documentation and prerequisites for AWS.

  • Fixed typo in cluster reboot instructions.

  • Added missing fields to API Package object.

  • Added note to say that the negation tag must be enclosed in double quotes.

  • Updated API Model Object documentation.

Apr 10, 2017

Cluster (2.6.1)

  • Upgrade Notes
    • No changes.

  • Cluster changes
    • Limit uploaded files in package scripts to 128MB. Anything larger should be done via a source url in the script. (Changed in 2.6.0.)

    • Fixed bug related to upgrade where bind from 2.6 job to 2.4.2 network-service fails.

    • Fixed bug where Service Gateways were too strict with the URI.

    • Fixed Postgres service create bug.

    • Fixed bugs related to Application Authentication Token failures.

    • Fixed potential leak of virtual network IP addresses.

    • Fixed display packages for a job when no environment variables are defined on that package.

    • Fixed bug resulting in IMs hanging during cluster restart process.

    • Added missing entries to bundled CA file.

    • Fixed bug with remove_splunk_forwarder.rb.

    • Fixed bug with failure to retrieve packages error after resource is swept in IM while the container is starting.

    • For IPAM in virtual networks, check if file exists before delete.

    • For IPAM in virtual networks, use UUIDs in filenames.

  • Chef changes
    • No changes.

Apcera Setup (2.2.4)

  • Updated VMware tools in Apcera-Setup images.

APC (2.6.1)

  • No changes.

Web Console (8.0.9)

  • Fixed issue with the Policy edit button not appearing.

  • Hard tags on jobs may be set (fixes regression).

  • Jobs that use packages that do not specify any environment variables will now list those packages on the "Environment" tab.

Orchestrator (1.0.0)

  • Added orchestrator-cli reboot support for cluster-object-storage.

Guide (2.6.6)

  • Updated sizing guidelines for MPD to require 3 central hosts for minimum HA.

  • Update Splunk documentation for connecting to cloud-based Splunk indexer.

  • Updated the OpenStack EE install docs for Liberty support.

  • Updated Sizing Guidelines for dedicated audilog-db host and proper component-db connectivity.

  • Updated the HTTPS documentation for multiple tlshosts.server_names and certificate update procedure.

  • Added policy variable version for the example ACME Policy use case.

  • Updated sizing guidelines for auditlog-database and added RPD diagram.

  • Updated policy variable migration example with proper policy variable names.

  • Fixed link in Keycloak configuration.

  • Updated the example for PUT /v1/policy with more explicit claims.

  • Updated sizing documentation for MPD/RPD that auditlog-db should be on didciated hosts; fixed typo.

  • Added note that policy claim value with the negation operator (~) must be in quotes.

  • Clarified the use of email.sendto for Zabbix alerts.

  • Fixed Azure typo in GCE install doc.

  • Converted Package.resource(s) to use allOf reference.

  • Added missing fields to Package object.

  • Added script for autodeploying docs.

  • Updated Splunk configuration documentation.

  • Updated job labeling docs for apps.

  • Updated blue/green deployment tutorial.

  • Clarified Docker support statement.

Feb 26, 2017

Cluster (2.6.0)

  • Upgrade Notes
    • Release 2.6.0 is a forward-only release. There is no support for rolling back to the release you are upgrading from.

    • The Linux kernel for cluster host images is upgraded to address After upgrading to release 2.6.0, please reboot all cluster hosts using the orchestrator-cli reboot command.

    • Be sure to use the orchestrator-cli reboot command to reboot cluster hosts. Use the --machines flag to reboot specific machines. See Automated Reboot Instructions. If you experience an issue rebooting a cluster host, please contact Apcera Support before attempting to reboot a cluster host manually.

    • VXLAN is the default protocol for virtual network tunneling. If you did not previously enable VXLAN in your cluster.conf, jobs in the virtural networks will be migrated to VXLAN and restarted. Support for GRE is deprecated and will be removed in the next major release. Note that to use VXLAN, you must open UDP port 4789 on each Instance Manager host.

    • The statsd-server component can be removed from cluster.conf. The Metrics Manager handles statsd traffic now and can be HA. Follow the appropriate Terraform procedure for removing components. Note that if you are running the statsd-server and metrics-manager on the same host, you must remove statsd-server.

  • Cluster changes
    • Added support to securely access Docker images from JFrog Artifactory registries.

    • Job scheduler is now tolerant to variable IM latency.

    • Added logic to ensure that IP addresses are properly reclaimed if the IM is not reachable exceeding the heartbeat interval but within IP refresh timeout (default 30 min).

    • Added logic for periodic purge of DNS records to clean up stale DNS entries, and do DNS publish for heartbeats to address lost publish messages.

    • Fixed bug where virtual network jobs could fail to restart after upgrade.

    • Crowd and Kerberos are no longer supported identity providers (deprecated in 2.4; removed in 2.6).

    • Added support to add root certificates for SSL certificate validation in semantic pipelines and service gateways. This allows users to add root certificate for semantic pipeline jobs and MySQL and Postgres service gateways to validate external servers, and for app-token semantic pipelines.

    • Added method for retrieving response of package upload to the backend via websocket to prevent timeouts with large packages.

    • Added logic to reannounce healthy application routes after disconnecting from NATS which may occur during network splits.

    • Changed the default logging format from cluster components for improved readability.

    • Removed the environment variable created from a service binding FQN in order to increase alignment with documentation.

    • Fixed bug to prevent duplicate environment variables being reported due to a binding.

    • Removed the ability to update job FQNs.

    • Updated the error message for Docker job creation handling when components are not healthy.

    • Disabled Nginx router requests buffering.

    • Replaced third-party statsd-server with Go based statsd-server as part of the Metrics Manager component.

    • Changed the default tunneling protocol for virtual networks from GRE to VXLAN.

    • Enabled HTTPS-only route designation. If enabled, HTTP requests will be redirected to HTTPS.

    • Implemented local IPAM that allows IP addresses to be allocated locally on the Instance Manager.

    • For policy, update events now contain the realms that were modified. Previously only added or deleted realms were reported on policy update events.

    • For policy, removed the non-functional OTP handling from the policy engine.

    • For the API, the binding name is no longer required when creating a binding and will default to BIND_.

  • Chef changes
    • Added chef role for deploying Consul. Reserved for future use.

    • Removed the installation of memcached from graphite servers.

    • Added the python-setuptools package to all cluster hosts, to allow installation of manual python packages as needed for debugging and system repair.

    • Allow an external auditlog database to be configured.

    • Enabled the local IPAM only if the enable_local_ipam is set to true and the Orchestrator version is 0.6.0 or higher.

    • Added options to terminate IM process using "force" or "evacuation" mode when rebooting an IM node.

    • Changed default reboot_delay to "now" from 2 minutes to speed up overall cluster reboot.

Apcera Setup (2.2.4)

  • Updated VirtualBox support to version 5.1.

  • Added multi-AZ support for AWS clusters.

  • Set reboot delay default to "now" instead of 2 minutes.

  • Set instance manager termination to "force" mode, which allows job evacuation to happen immediately when rebooting an instance manager machine.

  • Upgraded Orchestrator to 1.0.0.

  • Fixed printing of the vSphere password prompt.

APC (2.6.0)

  • Added —-cacerts option to service bind command to update a Certificate used for semantic pipelines.

  • Added —-cacerts option to provider register command to add Root Certificate for mysql and postgres providers.

  • Added —-cacerts option to job update command to add Root Certificate to job for app-token.

  • Fixed missing line break on errors being reported on package upload errors.

  • Fixed bug where zeroes would be trimmed during package upload progress.

  • Fixed bug that caused jobs created but failed to start (via the --start flag) from being properly removed as expected.

  • Removed flag for updating job FQN local names from apc app update command.

  • Allowed APC users to subscribe to policy document change events. The FQN for the subscription should be of the form policydoc::/policy or policydoc::/policy:: for add, update, or delete.

  • Clarified job status when using apc job list so that job instance health is considered and additional output given when in a warning state.

  • Added a --force flag to apc update to provide a means of downloading the version of APC from the cluster, regardless of your local version info.

  • Removed hidden non-functional apc otp commands.

  • Added --package flag to app/job creation which allows creating a job from an existing package in the cluster.

  • Updated the app from package command to accept the same flags as app create.

  • Fixed bug where packages may have been left in a failed state if apc package import failed.

  • Fixed bug that caused the -q, --silent flag to only enable batch mode without also silencing output.

  • Fixed bug that caused staging output to be abridged on Windows.

  • Changed default update behavior to require the current version of APC to exactly match the version reported by the API server.

  • Reduced time between apc version checks from 1 hour to 10 minutes.

  • Added command apc cluster usage to create cluster usage report.

  • Fixed bug that could cause APC to exit unexpectedly when given an invalid FQN.

Web Console (8.0.6)

  • Job environment page has been redesigned allowing Job environment variables to be edited and improving readability of package environment variables.

  • Root certificates for Semantic Pipelines may be added/removed/updated.

  • Job routes may be marked as HTTPS-only, which will automatically redirect traffic from HTTP (port 80) to HTTPS (port 443).

  • Only the features that a user is allowed to access, as defined by policy, are visible on the screen.

  • Generic and network service types added.

  • Persistent storage is now enabled by default for Minio services.

  • Docker launcher now allows users to choose whether to pull from the Docker registry or to use the cluster's cache.

  • Tooltip has been added to explain role of quotas on Job pages.

  • Mobile devices now show a different page header, improving usability.

  • Improved error messaging for actions denied by policy.

  • Improved formatting of policy update audit event details.

  • NFS mount path may be set when binding an NFS service.

  • Minor error messaging improvements.

  • Added client-side validation for certificates.

  • The package names on the "Environment" tab for a job now link to their appropriate packages.

  • When first navigating to the console, unauthorized requests will no longer be made and produce audit log noise.

  • Graph labels will no longer migrate over time in Firefox.

  • Allow-egress checkbox will maintain its state when navigating between pages.

  • Cursor is hidden from user when viewing policy (not editing).

  • Route metrics have been temporarily disabled until the API is stable.

  • Build ID now shows the correct information.

  • Jobs may be promoted to Service Gateways again without generating an error.

  • Error messages generated while uploading manifests will now display correctly.

  • Deleting packages from the packages list now correctly displays waiting indicator.

  • Custom params on services display correctly based on service type.

  • Job log will automatically scroll when appropriate.

  • Jobs listed in route list are now left-aligned.

  • When a "0" value is used for a port, it is displayed as "Chosen by system".

  • Tooltip added to lock icon for HTTPS-only routes.

  • Graph legend will stay fixed-in-place on all browsers.

  • Cancelling a file upload will not prevent additional file uploads.

  • The "Database" and "Persistence Provider" fields are optional when creating a service.

  • Certificate upload button has been removed from services that do not use a semantic pipeline.

  • Certificate upload button allows subsequent uploads.

  • Data tables with columns ending in multiple capital letters will display correctly.

  • Session and token invalidation is handled more-gracefully with greatly reduced API/audit noise.

  • The state of combo boxes will correctly reset when forms are dismissed.

  • Add-route form on job pages closes when navigating away.

  • The list of services will now update itself after deleting a service in a timely manner.

  • Job environment variable fields no longer disappear when navigating away and back.

  • Fixed bug so that logging out from the create-capsule form will not appear as if the API disconnected.

  • Fixed bug so that sign-out now works correctly for all pages when using Microsoft Edge.

  • Fixed bug so that labels on a job summary do not line-break.

  • Made the Submit button when editing job environment variables greyed-out until valid data has been entered.

  • Updated the Docker image launcher to work more smoothly with 3rd-party registries.

  • Added inline help text explaining that Docker Hub image names work without entering a full URL.

Orchestrator (1.0.0)

  • Updated the Orchestrator image to orchestrator (1.0.0) trusty; urgency=medium.

  • Added support for secret storage and credential prompts.

  • Added instance manager VM tracking.

  • Added YUM repository support.

  • Added skip-machines flag to reboot command.

  • Fixed AWS provisioner teardown.

  • Fixed issue with reading machine wait time after reboot.

Guide (2.6.4)

  • Added info about specifying UID/GID for app process.

  • Added API example for creating a service w/parameters.

  • Added Elasticsearch to known list of working Docker images.

  • Clarified API docs for package resource uploading.

  • Added APC CLI Reference documentation.

  • Deployed the documentation to using an --https-only route.

  • Added release note that 2.6.0 is a forward-only release; no rollback support.

  • Added /v1/cluster/admin API endpoint.

  • Added manifest expansion API endpoint.

  • Removed non-existent "weight" field from Job object.

  • Added topic for configuring the IM package cache pruning.

  • Made improvements to package script docs and Nodered tutorial.

  • Updated What's New and Release Notes.

  • Added manifest expansion API endpoint to the API documentation.

  • Reverted NFS changes.

  • Added route metric events to event reference.

  • Updated web console task for creating Docker jobs.

  • Updated the Terraform file structure for OpenStack.

  • Removed Crowd and Kerb support from docs (deprecated in 2.4; removed in 2.6).

  • Updated NFS documentation with bindfs usage.

  • Added docs for JFrog Artifactory Docker support 2.6.0.

  • Added https-only flag to multi-resource manifest docs.

  • Added information about enforcing HTTPS-only on jobs using Apcera-HTTPS header.

  • Added documentation for enabling local IPAM and deprecating GRE.

  • Added standalone topic for writing package scripts.

  • Added documentation for enabling HTTP router synchronization.

  • Added documentation for using custom CA certs for integrating with existing MySQL and Postgres databases for service creation and binding.

  • Added documentaton for implementing HTTPS-only routes.

  • Updated cluster configuration settings for router synchronization.

  • Updated sizing guidelines to remove statsd component (now handled by Metrics Manager).

  • Added documentation for using query->target_str to match route patterns.

  • Added example for adding multiple routes to a job.

  • Added example for using the --volumes flag with Docker images

  • Added documentation for using /TEMPLATES feature with Docker jobs.

  • Updated monitoring docs, including a typo fix in the apzabbix block of the monitoring configuration example.

  • Updated tutorials at

  • Updated Java New Relic tutorial and package script.

  • Updated CE install docs.

Jan 06, 2017

Cluster (2.4.2)

  • Upgrade Notes
    • If you have jobs affected by the virtual network issue fixed in this release (see below), after upgrading to release 2.4.2, stop all jobs in the virtual network, leave the virtual network for all jobs, join all jobs to the virtual network, and start all jobs in the virtual network.

  • Cluster changes
    • Fixed bug that could result in duplicate IP addresses for jobs in a virtual network.

  • Chef changes
    • No changes.

Apcera Setup (2.2.3)

  • Added fix for VirtualBox VM create failure.

  • Added fix for deployment failure in VMWare/VirtualBox.

APC (2.4.2)

  • No changes.

Web Console (6.0.11)

  • No changes.

Orchestrator (0.5.3)

  • No changes.

Guide (2.4.4)

  • Updated these release notes.

  • Added cluster reboot timing considerations.

  • Corrected units for disk/mem/netmin/netmax job resources in API docs.

Dec 22, 2016

Cluster (2.4.1)

  • Upgrade Notes
    • Updated kernel including a patch for CVE-2016-8655.

    • If you are uploading very large packages (more than 2GB), increase the staging_coordinator.memory.

  • Cluster changes
    • Patched the Linux kernel to address the recently discovered race condition that can be exploited to gain kernel code execution from unprivileged processes (CVE-2016-8655). See

    • Updated to Go version 1.7.4.

    • Fixed the 2.4.0 known issue related to very large package upload.

    • Added retry attempts to fix the 2.4.0 known issue related to the Postgres DB replication state check.

    • Updated the env_map template parameter with job process environment variables.

    • Deprecated the Crowd and Kerberos identity providers (they will be removed in a future release).

  • Chef changes
    • Fixed "eth0" hardcode regression bug for BareOS installations.

Apcera Setup (2.2.2)

  • Updated kernel including a patch for CVE-2016-8655.

  • Reboot required for users updating their clusters. Follow the on-screen instructions after a deploy to reboot.

  • Reboot required during new OpenStack installations. Follow the on-screen instructions after a deploy to reboot.

  • Added Ruby 2.2.5 package.

  • Added Python 2.7.12 package.

  • Added Python 3.5.2 package.

  • Added additional virtual disk for vSphere so it can be used for Instance Manager or Package Manager of Apcera Platform.

  • Added fix for proper error returning from vSphere resource lookup.

  • Added fix for parsing logic for finding vShpere objects.

APC (2.4.1)

  • Fixed issue by reusing less connections in the HTTP client from APC for more reliable transmission behavior.

Web Console (6.0.11)

  • No changes.

Orchestrator (0.5.3)

  • No changes.

Guide (2.4.3)

  • Added DRAFT documentation for using git-crypt to encrypt cluster config files.

  • Added deployment checklist for installing Apcera EE.

  • Updated the documentation for using the env_map parameter for runtime templating.

  • Strengthened and clarified the statement discouraging use of Basic Auth.

  • Updated AWS screenshots for CE installations using acpera-setup.

  • Removed auth header parameter from operations in the API documentation.

  • Added documentation for working with large package files.

  • Remove postgres_stateless from policy examples.

  • Added documentation for Component DB master failover and recovery.

  • Removed OTP documentation.

  • Clarified documentation that job FQNs are immutable.

  • Fixed minor typos.

Nov 11, 2016

Cluster (2.4.0)

  • Upgrade Notes
    • To install or upgrade to release 2.4.0, you must use Orchestrator 0.5.3 to take advantage of key new features.

    • To upgrade to release 2.4.0, you must upgrade from Apcera Platform 2.2.3 or 2.2.2. No other upgrade path is supported. If you upgrade from 2.2.2, you will need to reboot all cluster hosts due to a kernel change.

    • If upgrading, move the Auth Server from the Singleton host to the Central host to take advantage of new HA functionality. See Migrating to HA Components.

    • If upgrading, enable VxLAN in cluster.conf.erb and redeploy the cluster.

    • For virtual networks, OVS is upgraded from 2.4 to 2.5.1. If you are upgrading an existing cluster to 2.4.0, during the deploy you may experience (possibly multiple) brief network connectivity outages between jobs on a virtual network as each individual IM is updated.

    • For virtual networks, if you upgrade an existing cluster to 2.4, and then update to VXLAN, on the second deploy jobs in the virtual network will get restarted.

    • For virtual networks, if you are updating an existing cluster to 2.4 and you enable VXLAN, jobs which are part of virtual network will be restarted during the deploy, with possible periods where jobs cannot connect to each other as the deploy proceeds.

    • There is a known issue with the audit log for policy update; the policy.update log entry is missing an "update" JSON Patch payload.

    • There is a known issue using IPsec for cluster traffic encryption where IMs on different providers may result in NATS timeouts due to MTU discrepancies. IPsec is a BETA feature. If you plan to use IPsec, please contact Apcera Support.

    • If IPsec is enabled, reboot of the Postgres Master DB may fail due to a replication streaming check dealy. The workaround is to disable IPsec, reboot, and re-enable IPsec.

    • In certain networked environments where bandwidth is limited, you may receive an EOF or connection error during package uploads.

  • Cluster changes
    • Added policy soft invariance for writes so that the policy version is validated and stays the same throughout the update.

    • Improved audit logging by adding details to the audit logs generated from a job update, a job binding/unbinding to a service or another job, a job joining or leaving a network, routes being added to or removed from a job and a log drains being added to or removed from a job.

    • Added new audit event types for actions like, issuing a new authentication token to a user, job/capsule/app connect, stopping an instance of a job, snapshotting an instance of a job, adding/deleting log drains to/from a job and for promoting and demoting a job as a service gateway.

    • Added details to the policy updated, deleted, or added audit events to indicate the realms that changed and the policy document version number where that change went into affect. Also, policy document additions were being recorded as policy update events and this has been changed to be policy document add events.

    • Added resource access allowed and access denied events when a resource access attempt is initiated. These events contain the claim type and optionally the claim value that was checked to determine if the requester indeed had access to the resource. The logging of these events can be controlled by the the log claim type. Log allow on a resource will cause access allowed events to be written. Log deny will cause access denied events to be written. Log none will suppress log access denied or access allowed. The log claim associated with the most specific FQN will determine the claims in affect and log none takes precedence over low allow and log deny.

    • Added CPU Total metric to the instance resource usage events.

    • Allow job restart when mapped to TCP using the TCP auto realm. Added the autoprovision job attribute to record when a TCP route was mapped to a job using the TCP auto realm. ("on route::/tcp/auto { { permit map }}"). On restart check both the tcp realm for the job as well as the tcp/auto realm if the autoprovision flag is true.

    • New NFS providers will default to NFSv4 if no version is specified. Previously new NFS providers would default to NFSv3.

    • Added support for Amazon's Docker container Registry (ECR). The registry can be used via APC only and not Web Console in this release.

    • Extended the manifest deploy API to use substitution variables inside a manifest json file. Values for variables may be supplied via CLI or environment variables.

    • Extended the manifest deploy configuration to allow deletion of job links and service bindings. Also, the packages can now be updated on existing jobs during manifest re-deploys.

    • Recorded Docker image id in Docker image package.

    • Added the sp.disable job claim type which allows the policy writer to disable semantic pipeline to job binding by either service name, provider name, or protocol. Deprecates the service binding --disable-sp option.

    • Changed the payload field type in the audit log API response. Now payload will contain an array of JSON Patch ("") formatted elements to provide more details about the audit logs related to a job.

    • Added the string field Status to the "/v1/instance_managers/:uuid/instances" endpoint.

    • Added defaulting to latest tag on manifest deploy when using Docker images if not specified.

    • Added Docker image id to the output of apc package show.

    • Added username and password fields to the multi-resource manifest's Docker image declarations.

    • Added boolean pull field to the multi-resource manifest's Docker image declarations; if pull true it pulls the image's manifest from the registry always; if pull is false it tries to find and use the image manifest available locally in the namespace.

  • Chef changes
    • Added enhanced ldap-basic authentication configuration to deploy (cluster.conf). This exposes the LDAP Active Directory and custom field configuration feature.

    • Disabled swap on the Instance Managers.

    • Added code to handle Gluster snapshot disable/enable scenarios better, and force administrative action to remove snapshots if any exist before replacing a server.

Apcera Setup (2.2.1)

  • Added support to monitor your Apcera Platform cluster using Zabbix.

  • Added support for the AWS us-east-2 region.

  • Upgraded all images to kernel 3.13.0-83.127.apcera6.

  • Appended version to OpenStack images.

  • Added reboot command to safely reboot the cluster after an upgrade.

  • Added Minio server 2016-10-24 package.

  • Added Minio client 2016-10-07 package

  • Fix the compatability issue with macOS Sierra.

  • Fix a bug that could cause panic while terminating an Azure VM.

APC (2.4.0)

  • Added ability to create package using staging pipelines (apc package create). (Note that this command used to be an alias for apc package build but with this change that is no longer true.)

  • Added support for ".apcignore" file to allow developer to specify files to ignore when running apc app create, apc app deploy, or apc package create.

  • Fixed a bug which caused a running job to be stopped if a job link command failed due to policy checks.

  • Added apc route list which lists routes from all jobs. It has the same filter options as apc job list.

  • Fixed a bug which caused apc rule create to behave differently if the --batch flag was given, resulting in failure.

  • Modified apc audit --verbose to show the payload included with the event.

  • Fixed a bug which caused some output which had percent signs to become distorted.

  • Added --pull flag on apc docker run command to pull the latest manifest from the registry for the image specified in the command. Default behavior of the command is to not pull the image from the registry if available in local cache. In the past, apc docker run would always go to the registry.

  • Added extended flags to apc manifest deploy to supply values for substitution variables.

  • Fixed a bug which printed some results from apc event subscribe in scientific notation.

  • Fixed a bug which caused apc job logs to reprint log lines.

  • Clarified output during job show regarding job state.

  • Fixed a bug which caused some errors to not be handled properly, resulting in APC exiting unexpectedly.

  • Fixed an issue that caused apc login to not logout when logged in and --batch set.

  • Added output to indicate the amount of data being uploaded when uploading a package to give a better understanding of the upload speed.

  • Modified start, stop, and update methods (such as apc app start) to exit with 0 status code when the system is already in the requested state (for example, starting an app that is already running).

Web Console (6.0.11)

  • Fixed bug affecting multi-resource manifest file uploads.

  • Error messaging now displays when appropriate on multi-resource manifest uploader.

  • The correct value for platform build id is now shown in the help menu.

  • Audit event payload is shown in detail popup for policy access events.

  • Tables now show a pointer cursor when the row is clickable.

  • API Unavailable popup now shows a spinner while reconnecting.

  • API Unavailable popup now has a button for reloading the page.

  • KeyCloak authentication requests less information from the server; fixes a situation where running the KeyCloak server behind a proxy (such as when running it in the cluster) would generate errors

  • Audit log detail popup visual polish.

  • Audit log detail popup now links FQN to appropriate resource detail page.

  • Policy-in-context now supports FQN template wildcards.

  • Loading indicator no longer shows permanently on Capsule list page when the user doesn't have policy to view any packages.

  • Switching users no longer shows stale cached information from the previous user.

  • Versions of KeyCloak which require a nonce parameter will now accept authentication requests.

  • API Unavailable popup now shows correct link in Community Edition.

  • Label for user on audit log detail log window is consistent with other labels.

  • Auth audit event details are now visible.

  • Unauthorized users do not show in event detail window.

  • Services may now be unbound from jobs on the service detail page without error.

  • Instance manager count no longer reports as "0" while fetching data.

  • Test and build stability improvements.

  • Route URLs now display on the Route list page and on the Networking tab of Job pages.

  • This release just fixes a build error which prevented the previous RC from being built.

  • Added support for KeyCloak authentication.

  • Detailed information about audit log events can now be viewed by clicking on the log line.

  • Policy document name field width increased.

  • When binding a job to a service, a name for the binding will be auto-generated if not provided.

  • Service Gateways are now visible in "All" jobs list.

  • Task stream window will remain open after the task completes.

  • Line numbers and syntax highlighting are now shown when viewing policy documents.

  • Error message copy improvements.

  • Packages may now be deleted from the list view.

  • RabbitMQ, gnatsd, and Minio services may now be created.

  • Messaging is now shown to the user when the API becomes unreachable.

  • Stagers and staging pipelines may now be deleted.

  • Improved results when viewing the policy related to a specific resource.

  • State is now shown in the instance list for an individual instance manager.

  • Namespace selection is now persistent after deleting a job.

  • Discovery addresses for jobs may now be provided when joining a job to a network.

  • Policy documents may now be exported/saved as a file.

  • Policy data entry form will no longer get stuck in a loading state when the server returns an error.

  • Select and combo box popups no longer show up behind other elements.

  • Empty columns in policy data tables no longer generate errors.

  • Action icons on table rows will remain visible when there are multiple actions, regardless of viewport width.

  • Home page cluster name will now show a tooltip on hover showing the full cluster name.

  • Some units relabeled to reflect base-2 vs base-10 (i.e. 'MiB' vs 'MB' for 1024x bytes and 1000x bytes respectively).

  • Contextual help links for service creation form updated.

  • Google/Kerberos authentication will not generate errors when credentials have been entered and deleted.

  • Authentication popup will no longer be hidden behind graphics with small screen heights.

  • Long lines of text will no longer overflow the task stream window (for example, when launching Docker images).

  • Compatibility fixes for MS Edge.

  • Setting environment variables on apps created from Docker images no longer generates API errors.

  • Launching images from insecure (HTTP vs HTTPS) Docker registries no longer generates errors.

  • Adding service bindings to a job now generates the appropriate job link.

  • Job log lines are now displayed in the correct order.

  • "Updating" instances are no longer considered as unavailable, improving the accuracy of running instance counts.

  • Action buttons no longer get pushed off the screen when resources have a long namespace or local name.

  • The browser will now automatically reload the page when the cluster is upgraded.

  • "Advanced" fields now appear in the correct place on the service creation form when using custom parameters.

  • Signing out will now invalidate the session token server-side.

  • Jobs now more accurately reflect their health when they are starting.

  • Unknown API errors no longer display an empty message.

  • Added support for KeyCloak authentication.

  • Policy document name field width increased.

  • When binding a job to a service, a name for the binding will be auto-generated if not provided.

  • Service Gateways are now visible in "All" jobs list.

  • Task stream window will remain open after the task completes.

  • Policy data entry form will no longer get stuck in a loading state when the server returns an error.

  • Select and combo box popups no longer show up behind other elements.

  • Empty columns in policy data tables no longer generate errors.

  • Action icons on table rows will remain visible when there are multiple actions, regardless of viewport width.

  • Home page cluster name will now show a tooltip on hover showing the full cluster name.

  • Some units relabeled to reflect base-2 vs base-10 (i.e. 'MiB' vs 'MB' for 1024x bytes and 1000x bytes respectively).

  • Contextual help links for service creation form updated.

  • Google/Kerberos authentication will not generate errors when credentials have been entered and deleted.

  • Authentication popup will no longer be hidden behind graphics with small screen heights.

  • Long lines of text will no longer overflow the task stream window (for example, when launching Docker images).

  • Compatibility fixes for MS Edge.

  • Setting environment variables on apps created from Docker images no longer generates API errors.

  • Launching images from insecure (HTTP vs HTTPS) Docker registries no longer generates errors.

  • Adding service bindings to a job now generates the appropriate job link.

  • Job log lines are now displayed in the correct order.

  • Updating instances are no longer considered as unavailable, improving the accuracy of running instance counts.

  • Action buttons no longer get pushed off the screen when resources have a long namespace or local name.

Orchestrator (0.5.3)

  • Added Terraform modules for Azure and GCE. Note that currently you must use an external load balancer for these modules.

  • Added reboot command to facilitate cluster rolling restart.

  • Added reboot required warning messages at end of deploy run.

  • Added support to deploy remote release bundle.

  • Deprecated native provisioners support.

  • Added 'Reboot' status column to status command output.

  • Added backup and restore commands to support orchestrator VM migration.

  • Added native to generic provisioner conversion through repair command.

Guide (2.4.2)

  • Updated the BareOS intallation instructions.

  • Published GCE installation documentation using Terraform (new for 2.4.0).

  • Added info about conn_max_life_time setting for package manager.

  • Added note that apcera-setup requires direct internet connection.

  • Removed info about asynchronous service gateway APIs.

  • Updated Terraform installation instructions for specific supported versions.

  • Added post-installation tasks documentation for EE deployments.

  • Added documentation for Orchestrator 0.5.3, including rolling reboots and automated backup and restore.

  • Added documentation for Azure EE deployments.

  • Added docs for Amazon ECR registry support.

  • Updated manifest resource changes are handled.

  • Added docs for apc route list command.

  • Documented default namespace related claims.

  • Added documentation for disabling automatic semantic pipeline generation.

  • Updated OpenStack EE and CE installation docs for minor bugs.

  • Updated supported versions list for 2.4.

  • Published Apcera 100 Docker image support reference page.

  • Added docs for Amazon ECR registry support.

  • Added docs for Docker 'pull' flag in APC and MRM.

  • Added Keycloak auth provider documentation.

  • Added docs for Minio service gateway.

  • Document extendeded LDAP configuration for AD integration.

  • Updated API docs for deploying multi-resource manifests.

Oct 24, 2016

Cluster (2.2.3)

  • Upgrade Notes
    • Apcera recommends updating to 2.2.3 immediately to address the copy on write vulnerability known as Dirty COW. See the upgrade instructions.

    • On upgrade the kernel for each cluster host is patched. You must reboot each cluster host following the reboot procedure

    • You must update manually update the Orchestrator kernel using apt-get. Refer to the Orchestrator upgrade instructions.

  • Cluster changes
    • Updated the Linux kernel to patch the copy on write vulnerability (Dirty COW).

    • Fixed bug that prevented the creation of a network service with a duplicate name in a different namespace.

  • Chef changes
    • No changes.

Apcera Setup (2.2.0)

  • No changes.

APC (2.2.3)

  • No changes.

Web Console (4.1.3)

  • No changes.

Orchestrator (0.5.2)

  • No changes.

Guide (2.2.3)

  • Updated release notes.

  • Updated cluster and Orchestrator upgrade instructions.

Oct 14, 2016

Cluster (2.2.2)

  • Upgrade Notes
    • Upgraded from Go 1.6.3 to Go 1.7 to support macOS Sierra.

  • Cluster changes
    • Fixed bug where NGINX timeout makes user app fail.

    • Fixed bug so that slashes can be used in Docker registry passwords for MRM files and via the command line.

    • Fixed kernel upgrade bug.

    • Fixed bug that prevented read-only file system from being recognized by the Monitoring Server (Zabbix).

    • Fixed bug preventing Gluster snapshots from being properly disabled during cluster upgrade.

    • Fixed bug that caused Zabbix to fail to deploy if the component-database and auditlog-database were on the same host.

  • Chef changes
    • Fixed bug where read-only disks were not being picked up by the Monitoring Server (Zabbix).

    • Added support for enabling and managing HTTP Router tagging behavior.

Apcera Setup (2.2.0)

  • No changes.

APC (2.2.2)

  • Upgraded from Go 1.6.3 to Go 1.7 to support macOS Sierra.

Web Console (4.1.3)

  • Route URLs now display on the Route list page and on the Networking tab of Job pages.

Orchestrator (0.5.2)

  • No changes.

Guide (2.2.2)

  • Documented Staging RAM and size in cluster.conf.

  • Added reboot check for Riak and Gluster.

  • Fixed the policy seal! example.

  • Removed refs to generic services supporting semantic pipelines.

  • Changed provider to provisioner for the EE installation documentation.

  • Updated the docs to say that NFS providers default to NFSv4.

  • Added apc/docker run comparison notes.

  • Added link to cluster.conf docs to Dconf reference.

  • Added restart_mode to MRM doc.

  • Removed the "The Apcera API is subject to change without notice" note.

Sep 20, 2016

Cluster (2.2.1)

  • Upgrade Notes
    • This release provides support for Virtual Extensible LAN (VxLAN) to enable virtual networks instead of Generic Routing Encapsulation (GRE). Enterprise Edition customers must enable VxLAN in their cluster.conf and redeploy to use VxLAN (see Enabling VXLAN); Community Edition clusters are upgraded to use VxLAN, by default.

  • Cluster changes
    • Fixed nginx crash after route bind failures

    • Correctly copies default IP used for "AUTO" route provisioning.

    • Added cluster configuration option on Package Manager to limit the maximum life time for Postgres database connections.

    • Fixed issue extracting the namespace from an app token's job name.

    • Added support for specifying a virtual network's discovery address in a multi-resource manifest.

  • Chef changes
    • Added monitoring via Zabbix of the Postgres database replication status, which detects and alerts if the database replication fails.

    • Updated Splunk's default certificate to expire on 7/21/2016.

Apcera Setup (2.2.0)

  • Added Google Cloud Platform us-west1 region support.

  • Added rsnapshot 1.4.2 package to support rsnapshot based backups for NFS.

  • Added Apache 2.2.31 package.

  • Added Apache 2.4.23 package.

  • Added Nginx 1.11.3 package.

  • Added PHP 5.6.24 package.

  • Removed packages for Apache 2.2.29, Apache 2.4.18, Nginx 1.11.1, and PHP 5.6.19.

  • Fixed a bug which left the config.json file in an invalid state when a rescale failed.

  • Upgraded Orchestrator to 0.5.2.

  • Fixed a bug that caused deploy to fail if a user's SSH public key had a single quote.

  • Fixed a bug that caused SSH failures to return an unhelpful error message.

  • Fixed SSH support for Azure when no public key is given.

  • Increased OpenStack default volume size.

APC (2.2.1)

  • MacOS Sierra compatibility (preview).

Web Console (4.1.2)

  • Attempting to launch Docker Images will no longer generate a JavaScript error.

  • Service Gateways are now included in the "All" jobs list.

Orchestrator (0.5.2)

  • Fixed vSphere provisioner persistent data decoding problem.

  • Added repair command to reconfigure machines that have inconsistent state or regenerate missing client.pem key

  • Added status command to show cluster machine deploy status and list machine IP, name, hostname, ping status and tags, etc.

  • Added -machines and -tags flags to the deploy command to provide finer control on deploy command.

  • Added -update-name to force cluster update of the cluster name.

  • Added support for

Guide (2.2.1)

  • Added steps for enabling VxLAN in Enterprise Edition instead of GRE to create virtual networks.

  • Documented policy changes required for pulling Docker images (see Docker image whitelisting).

  • Clarifed HTTPS requirements for using app tokens; added app token mini-tutorial.

  • Fixed auth hostname for Kerberos-based authentication.

  • Clarified packages required if only using Docker-based workloads.

  • Added nameserver_rules to cluster.conf docs

Jul 20, 2016

Cluster (2.2.0)

  • Upgrade Notes
    • Check the Apcera Support Site for known issues in this release

    • For cluster authentication, the auth.<domain> API endpoint is deprecated in favor of the authentication-specific endpoint (<flex>.auth.<domain>). See the Authenication documentation.

    • To use Application Events, add the events-server to your cluster.conf. See the upgrade documentation.

    • You may need to manually add one or more flex-auth-server components to your cluster.conf. See the upgrade documentation.

    • The Apcera-provided images are updated with security updates. You will need to reboot each host after upgrade following the reboot procedure.

    • The standalone login dialog for signing in to the web console is deprecated. As a result, the redirect URLs are changed. If you are using Google Oauth, you will need to update your Google project.

    • The Health Manager and Metrics Manager can now be deployed redundantly. See the sizing documentation for details and the migration documentation for implementing.

  • Cluster changes
    • Cleanup Virtual Network State After Cluster Reboot.

    • Support IPSec across all nodes within a single Apcera cluster.

    • Fixed percentage based resource allocations in the instance manager configuration.

    • Changed Redis cleanup process to make multiple attempts at job log removal should first attempt fail.

    • Job log cleanup now has a small delay to ensure complete removal of entries.

    • Updated OpenSSL with security fix for CVE-2016-2106.

    • Fixed hard-coded MTU for veth- and vi- interfaces.

    • Provide better unique constraint violation error messages on job/record update.

    • Added support for job to job affinity tags in multi job manifest.

    • Fix for duplicated routes being added when using required ports in multi job manifest.

    • Fix for updating network max from running jobs in multi job manifest.

    • Fixed an issue with attempting to restore partially shut down containers on Instance Manager restart.

    • Fixed an issue with instance startup could error out cleaning up mounts that were already removed.

    • Fixed an issue with Instance Manager restart not always starting the process monitoring on existing instances.

    • Upgraded Nginx version to 1.10.1.

    • Fix IM to not send heartbeats or stats while it is disconnected from NATS.

    • Fix IM to properly release disk allocation when instance errors on creation before volume creation.

    • Fixed an issue with the Job Manager accepting the creation and update of a job with duplicate log drain URLs.

    • Running multiple Health Manager and Metrics Managers for HA is now both fully supported.

    • Restrict Gluster access to the known cluster subnets for security.

    • Allow jobs to be started/restarted for autoprovisioned routes using policy "on route::/tcp/auto" or "on route::/http/auto". Previously you would have had to permit map "on route::/tcp" or "on route::/http" or use the specific autoprovisioned address as part of the tcp or http namespace specification. For sites that do use the specific address no change is required.

    • Enhanced auth server to provide new Policy Schema REST API which allows metadata on policy claims and policy variables to be maintained on the server.

    • The /v1/oauth2/token endpoint now returns a 404 error.

    • Removed auth code from auth_server. The API endpoints for authentication through auth_server has been deprecated. The specification for "grant_type" is no longer required.

    • Fixed an issue where APC would seem to hang if a job failed to (re)start properly.

    • Policy now allows admins to launch Docker images by default.

    • Fixed an issue where job bindings would fail to validate that local names were set in its FQNs.

    • Removed policy preview handler.

    • Improved detection of existing network traffic control config on IM restart.

    • Move validation of broadcast/multicast routes to the job manager.

    • Added basic SG policy for RabbitMQ and MongoDB in the default cluster policy docs.

    • Fix multi-instance job linking.

    • Fixed an issue with Package Manager where S3 Delete operations would log as failed, but had succeeded.

    • Bump disk space granted to Docker layer downloader jobs from 1 GB to 2 GB.

    • Add permit start and stop enforcement for jobs.

    • Fixed an issue where the MySQL SP could sometimes cause CPU spiking on the IM host.

    • Removed auth code from auth_server. The API endpoints for authentication through auth_server has been deprecated. The specification for "grant_type" is no longer required.

    • Added apc manifest support for virtual networks.

    • Changed apc manifest routes to default to auto when endpoint not declared.

    • Added reporting position of JSON syntax error in payloads to API server.

    • Added user profile information to the response body of /token requests.

    • Changed staging jobs to timeout after running for over 1 hour.

    • Add google client id and scope info to auth /v1/info response.

    • Removed policy preview handler.

    • Added apc manifest support for broadcast/multicast routes in virtual networks.

    • Added support for job to job affinity tags in multi job manifest.

    • Added options to auth /v1/policy/::name for showing policy documents with variable references instantiated.

    • Add policy variable API to auth under /v1/pvar. API documented at "".

    • Enhanced the policy simulator to return all possible paths to a claim.

    • Enhanced the policy simulator to allow policy sim requestor to specify an epoch time for the base policy documents. The simulator rewinds policy documents back to that point in time.

    • Fixed the policy simulator to all filter on user defined claimtypes.

    • Enhanced the policy simulator to return the issuer as part of any INPUT claims reported.

    • Enhanced the policy simulator to be tolerant of double quotes in policy sim queries just like they appear in policy files.

    • Enhanced the APC policy sim command to allow the long value column to be split into multiple rows if the user chooses to do so.

    • Default Go package policy added. By default Go 1.6 will be used to satisfy a "go" dependency.

    • Fixed the policy simulator for incomplete results documented in Policy Simulator Results Advisory, as well as ensuring sim results are predictably ordered.

    • Enhance policy syntax to support system Policy Variables and Global Claims.

    • Fixed percentage based resource allocations in the instance manager configuration.

    • Updated OpenSSL with security fix for CVE-2016-2106.

    • Fixed an issue with attempting to restore partially shut down containers on Instance Manager restart.

    • Fixed an issue with Instance Manager restart not always starting the process monitoring on existing instances.

    • Fix IM to properly release disk allocation when instance errors on creation before volume creation.

    • Bump disk space granted to Docker layer downloader jobs from 1 GB to 2 GB.

    • Fixed an issue where the MySQL SP could sometimes cause CPU spiking on the IM host.

  • Chef changes
    • Removed auth code from auth_server. The API endpoints for authentication through auth_server has been deprecated. The specification for "grant_type" is no longer required.

    • Adding logging for when Riak needs to run a bitcask merge.

    • Prevent users from changing PM storage type on an already deployed cluster.

    • Changed logging level of the 'redis-server' component from 'verbose' to 'notice'.

    • Fixed a bug which prevented log rotation.

    • Fixed a bug which prevented multiple S3 Postgres backups from running on one host.

    • Fixed an issue preventing install of the platform on bare OS.

Apcera Setup (2.1.0)

  • Added Microsoft Azure support.

  • Added Google Compute Engine support.

  • Added an ssh command which allows the user to SSH into a machine by name.

  • Packages are now downloaded through a manifest instead of single archive.

  • Display Apcera Platform version information in status and info commands

  • Added OpenStack image service API v1 support.

  • Updated MySQL package to mysql-5.6.31.

  • Added MySQL Client 5.6.31 package to support MySQL backups.

  • Updated Postgres package to postgres-9.5.3.

  • Updated Go to 1.5.4 and 1.6.2.

  • Updated Node to 4.4.6, 5.12.0, and 6.2.1.

  • Updated Nginx to 1.11.1.

  • Updated the bootstrap flow to continue to import only intended packages while also not destroying your packages folder.

  • Fixes zend module loading error for PHP packages.

  • Fixed an NFS persistence bug with RabbitMQ package.

  • Fixed an invalid file path of passwordless sudo configuration.

  • Fixed a bug in the AWS Network Configuration to not default to previously chosen value.

  • Fixed an error in config command that caused the command to fail when the update check failed.

  • Fixed a bug that would allow invalid arguments and subcommands to be used without error.

  • Fixed the VM names are not consistent issue.

  • Fixed download logic so that if the file exists but does not match the expected size, it is redownloaded.

  • Fixed APC handling so that it would be redownloaded as necessary if deleted.

  • Fixed a bug in how new IMs were named when using the rescale command.

  • Fixed a bug that caused IMs created via rescale to not be configured with user provided private SSH keys.

  • Added ssh command that allows the user to SSH into a machine by name.

  • Added functionality to donwload packages using a manifest instead of single archive.

  • Display Apcera Platform version information in status and info commands.

  • Added OpenStack image service API v1 support.

  • Fixed a bug in the AWS Network Configuration to not default to previously chosen value.

  • Fixed an error in config command that caused the command to fail when the update check failed.

  • Fixed a bug that would allow invalid arguments and subcommands to be used without error.

  • Fixed the VM names are not consistent issue.

  • Fixed download logic so that if the file exists but does not match the expected size, it is redownloaded.

  • Fixed APC handling so that it would be redownloaded as necessary if deleted.

  • Fixed a bug in how new IMs were named when using the rescale command.

  • Fixed a bug that caused IMs created via rescale to not be configured with user provided private SSH keys.

  • Fixed stack name error for custom domains in AWS.

  • Updated check prompts on config.

  • Removed local file from failed release download.

  • Clarified apcera-setup update messages.

  • Added retry logic to package download.

APC (2.2.0)

  • Fixed bug that required users to manually enter non-empty start commands when using "app from package" instead of allowing the packages environment determine it if desired.

  • Added --disable-sp for preventing creating SP jobs on service bind.

  • Added UUID to the information shown for most objects (e.g. apc app show).

  • Fixed a bug in apc route add where the user's given path would be ignored in favor of an automatically generated default.

  • Changed the default http route added via apc route add to be automatically generated on the server.

  • Fixed a bug that reported that the root user may be locked even though apc Docker run clearly connected successfully.

  • Fixed a bug that would may have delete continuum.conf files during app create or app deploy.

  • Added dual stack support to http client for RFC 6555-compliant dialing.

  • Adjusted the way instances are listed for apps and jobs to be consistent with the console.

  • Added instance counts to console list.

  • If targeting a 2.2.X cluster with an older apc, trying to login to the cluster will trigger an autoupdate of apc.

  • Fixed a bug that may have caused the .apc file to be truncated on Windows.

  • Removed auth code from auth_server. The flag "–oldauth" has been removed.

  • Added reporting the position causing parsing errors in JSON format when using apc manifest.

  • Removed apc manifest start/stop commands.

  • Added apc docker pull and apc docker run --no-start commands; deprecated apc docker create command.

  • Removed APC policy preview command.

  • Added instance uuids to apc app logs stream.

  • Fixed a bug which may have caused connecting to an instance to panic.

  • Fixed a bug which may have caused app update to error if the app was running.

  • Enhanced apc policy show to support return of policy documents with Policy Variable and Global Claim references instantiated.

Web Console (4.0.2)

  • Improved copy on successful creation of app from Docker image.

  • Enable Google OAuth on localhost domain.

  • Set default time on audit log date range selector.

  • Fixed an issue where a user with an expired session token was prevented from logging in.

  • Fixed bug with Google OAuth on HTTP clusters.

  • Changed egress on a running job will now prompt to restart the job.

  • Added explanatory text has been added to the Multi Job Manifest uploader.

  • Added message so that users are warned when navigating away from a partially-filled form.

  • Added Docker image URLs so that they are now visible on job detail pages.

  • Updated labels on CPU graphs to be no longer rounded.

  • Fixed bug so that binding a job to a service no longer produces an error.

  • Removed a potential situation where a user may be prevented from logging in.

  • Made it so that the "Enable SSH" checkbox may not be unchecked for capsules.

  • Fixed tab-to-focus on Docker image launcher.

  • Fixed rendering error on bar graphs in Safari.

  • Improved the performance of inline-edit components.

  • Improved general rendering performance.

  • Fixed mouse hover interactions on network graph.

  • Improved job detail pages.

  • Enabled environment variables to be set when creating apps from Docker images.

  • Enabled egress at creation time for apps created from Docker images.

  • Improved the design of external links to Dockerhub from Docker image launcher design improvements.

  • Jobs with a persistent volume can now be restarted without error.

  • Fixed unintended compatibility warning message in IE11.

  • Jobs promoted to service gateways will now appear in the service gateway list.

  • When logging in, a user will no longer accidentally be routed to another most-recent route.

  • This also fixes bugs with page refreshes and deep-linking w/o being logged in.

  • Jobs are now restarted upon deleting service bindings.

  • Added button to display Docker form is now hidden when the form is visible.

Orchestrator (0.4.3)

  • No changes

Guide (2.2.0)

  • Updated Google OAuth documentation.

  • Added support for Acpera CE on Google Compute Engine.

  • Fixed issues with Papertrail/log drain tutorial.

  • Updated multi-job manifest docs with new features.

  • Added documentation for enabling cluster encryption using IPsec.

  • Updated the Ubuntu installation documentation.

  • Added SMB/CIFS documentation.

  • Updated Apcera Setup 2.1.0.

  • Added PKIX and updated cluster.conf documentation.

  • Added Policy Variables documentation.

  • Updated Google OAuth redirect URLs (for 2.2.0 and later versions).

  • Added documentation for apc docker pull and deprecate apc docker create.

  • Updated LDAP policy example (for all versions).

  • Added context-sensitive help redirect for creating policy variables (for 2.2.0 and later versions).

  • Added note about auto-generating HTTP route endpoints (for all versions).

  • Added IPsec documentation (for 2.2.0 and later versions).

  • Updated air gap installation steps with instructions to download the orchestrator-cli deb package (for all versions).

  • Updated title length for SEO (for all versions).

  • Added context-sensitive help redirect for creating policy variables.

Apr 14, 2016

Cluster (2.0.1)

  • Upgrade Notes
    • Chef no longer tries to access the internet to check for more recent versions of packaged/vendored gems.

  • Cluster changes
    • Fix audit log showing job events as updates when they were "start" or "stop".

    • Commented that the embedded ssh key is for transport to Apcera Platform, not for authentication or tunneling to jobs.

    • Fixed multi-instance job linking by randomizing links between instances.

    • Added iptables to restrict access to jobs in Apcera cluster to clients within the cluster subnet.

    • Don't attempt to stop jobs which aren't running with attached semantic pipelines.

    • API change - Disallow POST /unbind on running jobs to allow Egress off while job running from clients other than APC.

    • POLICY change - Default policy in docker.pol is enabled to allow "admin" users to pull Docker images from Docker Hub. Only applies to new deployments."

  • Chef changes
    • Chef no longer tries to access the internet to check for more recent versions of packaged/vendored gems.

Apcera Setup (2.0.3)

  • Upgraded Orchestrator to latest version.

  • Upgraded images to latest version.

  • Improved bootstrap speed for importing packages.

  • Added SHA to apcera-setup --version.

  • Improved prompt for Google users input.

  • Enabled gnatsd, S3 and RabbitMQ service gateways by default.

  • Fixed a problem where quitting mid-install may create orphaned VMs.

  • Fixed a concurrent issue when multiple goroutines to access IP cache.

  • Improved user experience of prompt messages in config command.

  • Implemented retry destroy on AWS if at first it fails.

  • Added logic to download cluster logs during install if the deployment fails.

  • Set default authentication provider to "basic" if Google is not enabled.

  • Fixed MySQL provider bootstrap failure in some race condition.

  • Modified APC commands to stream output into the logs for better feedback and debugging.

  • Clarified DNS and OAuth2 options.

  • Modified creation of MySQL and Postgres providers to avoid reimporting to provide better feedback.

  • Fixed wrong error message when VBoxManage is missing.

  • Improved SSH error messages printing in the log.

  • Added prompt to confirm acceptance of agreement.

  • Added optional auto VPC, subnet, and security group configuration.

  • Allow users to add google email addresses.

  • Added timeout counter for DNS changes.

  • Creates MySql and Postgres providers during bootstrap.

  • Added support to provide feedback during deploy.

  • Added retry for DNS API call.

  • Added support to automatically upgrade the orchestrator.

  • Added prompt to ask user to trust cert during config steps.

  • Modified config to, by default, reuse a previously generated HTTPS certificate.

  • Print warning about sensitive credentials in config.json.

  • Changed default resource to 2 CPU and 8GB RAM for vSphere.

  • Increased the dial timeout on waiting for the web console available, also changed timeout error to non-fatal.

  • Increased provisioning timeout to 90s.

  • Added Developer Portal link to deploy output.

  • Removed provider type printout.

  • Removed im-vg volume groups from OpenStack IMs and Storage Nodes.

  • Fixed the EOF error in DNS API.

  • Fixed issue causing redeploys to fail under certain conditions.

  • Fixed storage node configuration for OpenStack.

  • Updated VirtualBox setting VM names during creation.

  • Wait for instance to exist for AWS provisioning.

APC (2.0.1)

  • Fixed error message when v2.0 APC connects to 508 or older cluster.

  • Improved confirmation messages on multiple commands by displaying full FQNs rather than only local names.

  • Fixed an error in which Golang apps would not get the proper GOPROJECT set via app deploy.

  • Improved error messaging during filecopy and pipe commands.

Web Console (2.3.2)

  • Added curated list of Docker images has been added to the Docker launcher in Community Edition.

  • Updated Docker launcher to be a multi-step form.

  • Updated error messages to now display correctly in Docker launcher.

  • Changed the style of capsule OS catalog to more-closely resemble the new Docker launcher.

  • Fixed bug so that saving a policy document will no longer timeout if there is an error in the policy change.

  • Disabled the Next button when appropriate in the Docker launcher.

  • Fixed JavaScript error in policy preview.

  • Altered "My Sandbox" link on home page to send user to the all-jobs list screen.

  • Removed checkbox for toggling ssh on/off removed from capsule pages.

  • Fixed performance problem caused by leaking event handlers in the combo box component.

  • Hid capsule create button when the form is visible, making it consistent with how other forms work.

  • Hid the sidebar automatically after selecting an item at screen widths where the sidebar is toggleable.

  • Changed the cursor position by clicking with the mouse within the policy editor now behaves as expected.

  • Changed cluster RAM/Disk graphs to no longer show inaccurate dips in graphs due to the timing of reporting.

  • Changed links to cluster pages so they are no longer visible to users that do not have permission to visit those pages.

  • Updated error messages while creating/editing policy to now display correctly.

  • Changed copy on sidebar New button has been to Launch.

  • Reorded Launch menu items and updated the copy.

  • Reduced vertical padding on job log lines.

  • Updated job logs so that the leading whitespace is now preserved.

  • Fixed memory leak caused by broken event handler cleanup.

Orchestrator (0.4.3)

  • No changes.

Guide (2.0.6)

  • Rewrote sizing guidelines for Enterprise Edition.

  • Added steps for updating stager jobs.

  • Added redirects to nginx.conf for context sensitive help.

  • Removed deprecated NFS SG service IP parameter.

  • Updated console tasks docs, removed netflix reference.

  • Updated Apcera EE Installation on Ubuntu 14.04 (Bare OS) documentation.

  • Added a new topic for monitoring and managing job health.

  • Updated apc cluster commands.

  • Updated authorized redirect URIs for Google OAuth.

  • Fixed errors in service gateway API overview documentation.

  • Fixed broken links.

  • Removed single-tier service gateway type description.

  • Removed Riak migration steps (changing the package storage type of an existing cluster is not supported).

  • Added cluster.conf instructions for enabling NFSv4.

  • Updated external monitoring documentation.

  • Updated vSphere requirements for Apcera CE.

  • Fixed typos.

  • Added steps for updating stager jobs.

  • Updated snapshot API.

  • Added documentation for apcera-setup update service.

  • Updated documentation for automatic staging pipeline detection rules.

  • Fixed typos in virtual networking documentation.

  • Added bare OS installation instructions (BETA documentation).

  • Updated policy documentation for permit start/stop on job realm.

  • Added comparison between Docker and Apcera jobs.

  • Added bare OS installation instructions.

  • Updated OpenStack installation instructions using Terraform.

  • Improved search performance.

  • Updated port list for NFS and HA NFS.

  • Added API Explorer app/src with instructions for running it locally.

  • Moved doc search server side and consolidated to single page.

  • Fixed basicauth endpoint in auth documentation.

  • Reverted change that removed the Aardwolf release name.

  • Updated web console docs/screenshots.

  • Added Package.dependencies/provides to API models.

  • Added Alexa web tracking code to header.

  • Added policy simulator API docs.

  • Clarified use of new "flex auth" endpoints.

  • Changed config example on how to configure Riak GC settings.

  • Clarify statement around connecting via links to multiple instances.

  • Changed Javascript to bind search loading only when search box is entered (improves doc load speed and performance).

  • Added note about listing available software releases using Orchestrator CLI.

  • Updated instructions for creating Semantic Pipeline Rules.

  • Removed erroneous references to Kurma container format.

  • Added detail about Job.processes to API docs.

  • Removed reference to Ardwolf internal release name.

  • Added vSphere example to Gluster configuration documentation.

  • Added note that cluster-name cannot have the dot/period character.

  • Added continuum_max_fails and continuum_fail_timeout to cluster config docs.

  • Updated upgrade documentation with steps for removing components and making component config changes.

  • Updated hybrid cloud datacenter tag usage and examples.

  • Updated Java New Relic tutorial.

  • Updated S3 services documentation.

  • Added troubleshooting tips to component reboot instructions.

  • Updated Apcera CE documentation for the official launch.

  • Updated Docker persistence examples.

  • Fixed typo in apc pipe usage.

  • Updated Orchestrator documentation (removed older releases).

  • Added step to make Ruby Stager API example work.

  • Added note to S3 services about safe to ignore warning.

  • Updated Mongo, Redis, Memcache, RabbitMQ, and NATS services documentation.

  • Removed installation steps for OpenStack provider (deprecated) in favor of Terraform generic installation for OpenStack.

  • Updated OpenStack CE and EE installation steps.

  • Updated AWS EE installation steps.

  • Updated CE bootstrapping doc.

  • Updated the upgrade procedure documentation.

  • Updated google analytics tracking id.

  • Removed outdated Docker staging pipeline reference.

  • Removed reference to apc docker pull (not part of LTS1).

  • Updated the backup and restore documentation and removed DRAFT status.

  • Updated example manifest original max_size from 20 for log drains was too small for most cases.

  • Updated documentation for sample package ZIP file.

  • Updated google redirects to represent new oauth flow.

  • Updated APCFS examples to use CE default namespacing.

  • Update google redirects to represent new oauth flow.

  • Updated default_provider is now required.

  • Added defaultNamespacePrefix to namespace policy example and permissions documentation.

  • Updated the "MySQL from Docker with persistence" example.

Mar 29, 2016

Cluster (2.0.0)

  • Upgrade Notes
    • Apcera Platform is re-versioned to 2.0.0.

    • Apcera recommends that you upgrade to 2.0.0 as soon as possible during a maintenance window of your choice.

    • Note that 2.0.0 is a forward-only release. Once you upgrade to 2.0.0, you cannot rollback to a previous release.

    • Orchestrator version 0.4.3 is required to install Apcera Platform version 2.0.0.

    • To upgrade to 2.0.0, you must first upgrade to 508.1.5.

    • The default_provider is now required in cluster.conf. (Article)

    • If present, you must remove metrics_exporter from the stagehand plugins list in cluster.conf or the deploy will fail.

  • Cluster changes
    • Updated to Ubuntu Linux Kernel 3.13.0-83.127 plus an isolated patch for Open vSwitch.

    • Fixed security concern with virtual networking by adding src IP based firewalling.

    • Updated DConf parser to provide better error messages.

    • Added support for Multicast and Broadcast for Jobs using Virtual Networks.

    • The internal NFS server deployed for new clusters now supports NFSv3 and NFSv4. NFSv4 is required to deploy Docker containers that require persistent NFS storage.

    • Added Enterprise Edition support for installing highly-available, distributed NFS server based on GlusterFS and GaneshaNFS.

    • Added flexible-authentication server components. Existing auth-server remains unchanged, but its auth-endpoint will be removed in the future.

    • Deprecated authentication API endpoint "auth.domain", for example, in favor of new flex-auth endpoints specific to auth method used "new-endpoint.domain", for example,

    • Added error messaged that a Docker image may not exist if the registry returns a 401.

    • Added support for content-type json in the token handler.

    • Fixed binding "IP:Port/Hostname" collisions by rewriting port with a new available port in the IM.

    • Added support for the --net flag to the apc docker run command.

    • Fixed dangling job references in virtual networks when jobs are deleted via web console or other clients by moving relevant functionality from APC to the job manager.

    • Added support for multi-job manifest http "endpoint:auto" declarations mapping default routes.

    • Changed multi-job manifest deploy command so that a re-deploy restarts jobs with modified resources, ssh, process environment settings.

    • Updated default package resolution to use patched glibc Ubuntu 14.04 package.

    • Added LDAP extra sanitization for username.

    • Fixed TLS certificate validation for LDAP and correct URI handling.

    • Added container restore functionality for virtual networks. The IfEntry object is now restored upon IM restart. Also cleaned up libnetwork to work with restore.

    • Fixed apc login so that you are not prompted to logout if your token is invalid.

    • Check that instance volume exists before attempting to destroy it.

    • Fixed clean up of container directories.

    • Added logic to handle auto HTTP routes on job update.

    • Fixed bug that would incorrectly resolve any auto TCP routes present on job create.

    • Added binding for Cluster Monitor to static port 6768 to enable Zabbix monitoring of /healthz.

    • Added binding for Metrics Manager to static port 6767 to enable Zabbix monitoring of /healthz.

    • Fixed password authentication for Linux APC; preventing erroneous output.

    • Fixed route format in HM's remove target message to router.

    • Added Policy Simulator to perform simulations against policy using APC commands or REST API.

    • Fixed 'apc policy on' cmd by allowing '*' as a resource type in fqn queries.

    • Fixed instance metric reporting issue which resulted in occasional spikes seen in instance metric graphs.

    • Added retry attempts to stop stale staging coordinator jobs in order to trigger cleanup up finished staging coordinators.

    • Randomizing route host-side port selection for jobs.

    • Removed statsd from the instance metric accumulation pipeline, resolving random spikes in both instance reservation in usage.

    • Added logic to remove errored-state staging coordinators when the packages are marked failed.

    • Fixed Cluster Monitor "deadlock" hanging.

  • Chef changes
    • Changed the kernel to 3.13.0-83.127.apcera5.

    • Changed configuration to reserve 256MiB of memory on each Instance Manager for system services.

    • Updated configuration to allow disabling apceraOperations policy.

    • Added support for flexible-auth components.

    • Upgraded Graphite to 0.9.15.

    • Added monitoring test for corrupt data in the packages database.

    • Added proxy_next_upstream error; to router.conf

    • Enable Zabbix monitoring for the cluster-monitor via healthz.

    • Remove stagehand plugin for metrics exporter job.

    • Enable Zabbix monitoring for the metrics-manager via healthz.

Apcera Setup (1.5.3)

  • Minor bug fixes.

APC (2.0.0)

Web Console (2.2.8)

  • Implemented all-new look and feel for the entire web console.

  • Enabled easy deployment of Docker images and capsules using the web console.

  • Enabled easy app creation by uploading multi-job manifest file using web console.

  • Added UI for promoting jobs to service gateways.

  • Added link to download apc for 32-bit Windows.

  • Added home page with recently-used items, shortcuts, and links to tutorials.

  • Fixed OS selection button on capsule create for IE.

  • Removed preview step when saving policy documents.

  • Updated display so that the system build ID is now visible in help menu.

  • Made more clear the messages displayed during policy preview.

  • Updated the Support link in the help menu.

  • Updated capsule form to animate correctly.

  • Fixed bug so that signing out does not prevent signing back in.

  • Changed display so that the first field in an inline form is now automatically focused when the form becomes visible.

  • Changed capsule create form to indicated when no OS packages are available on the cluster.

  • Changed background color of login dialog error message to white.

  • Added right margin to Login dialog error message.

  • Improved error message when no authentication info has been provided.

  • Home page displays different content depending on targeted Apcera Platform edition.

  • Updated combo box dropdowns to now appear in the correct location on MS Edge browser.

  • Updated data tables to display consistently across browsers.

  • Made namespace navigator design improvements.

  • Updated Policy upload button to work for all supported browsers.

  • Made consistent the spacing of elements in subheader across all supported browsers.

  • Added progress spinner is now shown when deleting an item from a table.

  • Hid job UUIDs from the user.

  • Removed redundant display of job's type from the job summary page.

  • Removed instance count displayed between job/instance resource tables.

  • Removed redundant display of resource table on the job summary.

  • Change the element used to add a new key-value pair in lists to a link from a button.

  • Restyled the "New" button in main nav.

  • Made the copy in the "New" menu is now consistent with itself.

  • Made the key/value lists now fill the width of their container.

  • Fixed error message shown when attempting to delete a network that still contains jobs.

  • Fixed it so that adding env vars to a package now works as intended.

  • Made it so that the cluster version is now visible again in the help menu.

  • Fixed it so that button labels now display again in Firefox.

  • Fixed misleading error message shown when authentication fails.

  • Fixed sidebar footer to displays correctly in IE.

  • Restored elements that show on mouse hover over graphs.

  • Fixed labels in the sidebar to display correctly in MS Edge.

  • Fixed form controls to overflow correctly.

  • Added link to peer support for CE edition.

  • Fixed checkboxes to display consistent sizes across browsers.

  • Fixed horizontal scrollbars in the main content area for all supported browsers.

  • Positioned correctly modal dialogs on IE.

  • Removed JavaScript error which would show up in the dev console when viewing instances.

  • Added filtering for Jobs by label on the various job lists.

  • Added more-appropriate text for form-field labels for job labels and environment variables.

  • Added display for IPv4 addresses of connected jobs where networks are visible.

  • Updated behavior so that the "Enter" key now submits form on multi-job-manifest uploader.

  • Updated package environment variables on the job detail page to link to the appropriate package page where they may be edited.

  • Made copy adjustment for the docker-create button.

  • Made spacing adjustments throughout.

  • Made color adjustments to data center bar meter.

  • Made whitespace adjustents on the task-stream dialog.

  • Updated display so that all links now underline on hover.

  • Made design enhancements for the sidebar footer.

  • Made design enhancements for the policy editor.

  • Made design enhancements for the multi-job manifest uploader.

  • Updated commands and environment variables to displaye in monospace font.

  • Made "Ignore Volume" the default option in the "Volume Provider" dropdown on the "Create App from Docker Image" form.

  • Updated custom parameters to be configured for a service prior to it being created.

  • Added component to increase visibility of some inline-editable values.

  • Added link to download 32-bit Windows apc binary.

  • Made design enhancements for network association table on job detail.

  • Added functionality to preserve and display file extension when uploading a multi-job manifest.

  • Updated display to correctly show errors that occur while updating job resources or uploading manifests.

  • Updated the "No Data" text on line graph empty-state to appear in the correct position.

  • Fixed transient alignment problem when editing inline-editable values.

  • Removed unwanted wrapping of content in page header on narrow monitor widths.

  • Removed error that displayed when attempting to remove a job from a route.

  • Restored reservation/warning/quota lines on resource graphs.

  • Restored narrow time window options to the resource graph time-picker.

  • Removed Ubuntu 13 base capsule type.

  • Added logic to prefill forms that require a namespace with the user's default namespace.

  • Updated namespace navigator to sorte namespaces alphabetically.

  • Added functionality to display error message when attempting to start a job fails.

  • Added functionality to pre-expand the currently selected namespace.

  • Removed service gateways from job list.

  • Made form validation appear in-line without changing focus.

  • Made editable job start and stop timeouts.

  • Improved UI for communicating errors in policy.

  • Updated job FQNs in policy to now link to their respective jobs.

  • Added button to export policy violations to a CSV file.

  • Updated display so that data center instance counts are now represented as a bar graph.

  • Made services filterable by namespace.

  • Improved messaging for communicating unsupported user agents.

  • Removed invalid namespaces from audit log filter.

  • Combo boxes preserve their content when the keyboard focus changes

  • Fixed table sorting behavior on app and capsule pages.

  • Removed unwanted throbber from the ports list on job pages.

  • Updated many navigation elements to display a menu on right-click.

  • Reduced number of API requests on many pages.

  • Updated audit log display so that server errors are now displayed.

  • Corrected incorrect units for CPU quota on job resource card.

  • Made visible the error messages on network page.

  • Fixed JavaScript error in policy preview.

Orchestrator (0.4.3)

  • Forced upgrade. You must use Orchestrator version 0.4.3 to deploy Apcera Platform 2.0.0.

  • Changed orchestrator-cli version command to to include a short hash in the description.

  • Fixed bug in deploy command when neither --release nor --update-latest-release flag is given. (0.4.2)

  • Added logic to error out when no database initialized, or version missing from database. (0.4.1)

  • Changed the --update-latest-release command to now only update automatically to newer point releases. (0.4.1)

  • Added reattempts logic for orchestrator-agent connections. (0.4.1)

  • Added releaseinfo subcommand to display available releases. (0.4.1)

Guide (2.0.0)

  • Added tutorials for web console home page and developer portal.

  • Updated Walkthrough tutorial and screenshots.

  • Updated Apcera CE installation steps and requirements for vSphere.

  • Updated Apcera CE managing and troubleshooting sections.

  • Updated release notes for Apcera CE 1.5.x and web console 2.2.x.

  • Added PostgreSQL provider requirements and troubleshooting documentation.

  • Updated optional parameters for service provider registration.

  • Clarified riak settings example.

  • Added release notes for LTS 2.0.0.

  • Bumped version of docs to 2.0.0 to sync with LTS.

  • Added NFS server upgrade options.

  • Updated the Zabbix monitoring documentation to not alert on swap if there's plenty of memory left.

  • Added documentation for backing up MySQL database and connecting to an existing MySQL DB.

  • Added instructions and tutorial for migrating a MySQL database.

  • Added instructions and tutorials for backing up a MySQL database.

  • Made MJM toc entry consistent with page title.

  • Updated the API documentation for job bindings, and added documentation for using --trace mode.

  • Fixed API docs for PUT'ing package resources.

  • Adding Metrics Manager documentation to Ops > Storage with reference in vCloud installation docs.

  • Added note to Zabbix monitoring section about entities that will be overwritten by cluster.conf on redeploy.

  • Added instruction on how to add policy description to policy document.

  • Updated the APCFS and NFS documentation.

  • Added instructions for connecting to an external MySQL server.

  • Updated Installation Intro section.

  • Updated Apcera CE install steps for 2.0.0.

  • Document the use of the Apcera MySQL package.

  • Updated PostgreSQL service creation documentation.

  • Updated the docs for job scheduling and job affinity tags with explanation that job restart is required.

  • Updated Using the Web Console and Web Console task list documentation.

  • Updated the apcera-setup CE documentation for AWS.

  • Clarified filecopy documentation.

  • Updated NFS HA documentation for Gluster.

  • Documented the use of 'persistence_provider' option for mongo, rabbit, redis.

  • Updated job scheduling and tagging documentation.

  • Added new 404 page.

  • Added documentation configuring "postfix" option for mail relay agent alternative for Zabbix alerts.

  • Updated Orchestrator documentation for version 0.4.1 required for release 2.0.0.

  • Updated AWS requirements and configuration documentation for installing CE.

  • Updated network discovery documentation for Virtual Networks.

  • Fixed broken links.

  • Added note about how to find your s3 endpoint for AWS Terraform installations.

  • Added docs about new auth endpoints and deprecation of legacy auth endoint.

  • Fixed Gluster architecture diagram.

  • Removed version banner from home page.

  • Updated documentation for apcera-setup 1.4 local installs (Fusion, VirtualBox).

  • Added redirects for /faq and absolute URLs for and the changelog.

  • Updated readme with link checking info.

  • Added broadcast/multicast virtual network docs.

  • Added auto HTTP route documentation.

  • Added documentation for configuring HA NFS using Gluster.

  • Updated documentation for apcera-setup for AWS.

  • Added documentation for Policy Simulator.

  • Added documentation for new UI tasks.

  • Added documentation for service discover of virtual networks.

Feb 25, 2016

Cluster (508.1.5)

  • Upgrade Notes
    • Apcera recommends that you upgrade to 508 as soon as possible during a maintenance window of your choice.

    • Note that 508 is a forward-only release. Once you upgrade to 508, you cannot rollback to a previous release.

    • To upgrade to 508, you must first upgrade to 506.1.0.

    • Restart cluster components after upgrading to 508. See the reboot documentation.

  • Cluster changes
    • Update glibc in release bundle to resolve CVE-2015-7547 (requires component restart).

    • Prevent user jobs from sharing traffic with cluster components like the API server or Auth server.

    • Added logic to prevent HTTP router from retrying indefinitely.

    • Enabled disabling of debug ports on an Apcera component using the setting "webadmin.enabled=false" in the config file.

    • Removed restart-attempts from APC, the API, and internals (the field was previously deprecated).

    • Updated the Policy REST API to return the policy description field in the HTTP response block for policy name requests.

    • Fixed bug to calculate instance uptime rather than Instance Manager uptime from the cluster IM stats API.

  • Chef changes
    • Added option to disable debug ports.

    • Added AWS VPN endpoint.

    • Enabled Riak and Riak-CS Garbage Collection settings to be configurable.

Apcera Setup (1.3.2)

  • No changes.

APC (0.28.2)

  • You must use APC version 0.28.2 with Apcera Platform version 508.

  • Fixed broken Crowd and LDAP authentication on ​Windows APC client.

  • Added --unlock-root flag to apc docker run to allow automatic unlocking of root user accounts in Docker images, re-enabling SSH.

  • Added manifest start, manifest stop, manifest status commands.

Web Console (1.20.4)

  • No changes.

Orchestrator (0.4.0)

  • Updated lexer for DConf files.

  • Added 'clusterinfo' subcommand.

  • Fixed vsphere provider to allow datacenter to exist inside of a folder.

Guide (

  • Updated Cluster Monitor ports.

  • Updated APCFS documenation to support for NSFv4 for LTS.

  • Updated Apcera CE documentation with steps to trust certs on Windows, plus other refactoring.

  • Added documentation for implementing NATS services using NATS service gateway.

  • Added documentation for configuring the Stagehand block in cluster.conf for service gateways.

  • Updated Apcera Setup installation steps.

  • Added static port for metrics manager.

  • Fixed invalid TOC item in navigation for Apcera Setup.

  • Updated conditional logic to support cluster and public versions. See README for details.

  • Updated installing new image documentation.

  • Updated these release notes for release 508.

  • Added documentation for configuring the cluster to support JESSIONID.

  • Changed occurrences of "Continuum" to "Apcera."

  • Updated some meta info in the repository so guide can be built as part of a release.

  • Fixed policy permissions section of the docs for Docker and Package Resolution claims.

  • Fixed typo in Namespace policy example.

  • Added definition for Stagehand to glossary.

  • Added documentation for configuring IP Manager.

  • Moved Static Site tutorial to be the first tutorial.

  • Reverted change to instructions for VirtualBox setup for network bridge.

  • Added link for trusting SSL cert on Linux.

  • Clarified which SPs support ephemeral credentials.

  • Added documentation for using TLS with Crowd.

  • Added documentation for Riak garbage collection.

  • Added missing architecture link.

  • Hid apcera-setup upgrade steps.

  • Added documentation for importing new images.

  • Added Sumo Logic integration tutorial.

  • Fixed deploy command for 504 upgrade.

  • Fixed broken links and typos.

  • Changed banner version to 508.

  • Corrected syntax on LDAP group policy example.

  • Updated default route suffix name policy example.

  • Added description of users required for providers and services.

  • Added example for routing rule to opt in and allow an IP and exclude all others.

  • Clarified that a unique user is required for each S3 service.

  • Documented dconf syntax.

  • Updated apcera-setup and Apcera CE installation instructions.

Feb 11, 2016

Cluster (506.1.0)

  • Upgrade Notes
    • Note that 506 is a forward-only release. Once you upgrade to 506, you cannot rollback to a previous release.

    • To upgrade to 506, you must first upgrade to 504.1.8.

    • Note that in 506 PostgreSQL is upgraded to 9.4. See Chef section below.

  • Cluster changes
    • Fixed bug with mapping multiple routes when using manifest.

    • Fixed beginsWith and endsWith policy operators to accept non-strings as right-hand argument.

    • Fixed bug in manifest deploy command related to updates of scheduling tags.

    • Fixed bug in the Instance Manager where bindings with a pipeline associated with them do not have a reference to that pipeline.

    • Added APC manifest support for package_env.

    • Fixed bug so that failed placement for jobs (i.e., due to IM unavailability) will no longer mark jobs as errored, unless they are configured to never restart instances on exit.

    • Fixed download/extracting packages timeout by adding retry flag error.

    • Changed manifest deploy to emit more task events.

  • Chef changes
    • Added option to customize or unset support URL.

    • Upgraded from PostgreSQL 9.3 to 9.4. Existing clusters will upgrade all internal database servers to PostgreSQL 9.4, and new clusters will be installed with 9.4.

    • Note that upgrade involves downtime for the component-database-master server, and thus the API, web console, and other management functions will be inoperable during the upgrade.

    • Added gnatsd to the default set of gateways configured by stagehand.

    • Fixed an issue with monitoring configuration for the Redis log server that prevented monitoring from working after a single-pass deployment on a new cluster.

Apcera Setup (1.3.2)

  • No changes.

APC (0.28.1)

  • You must use APC version 0.28.1 with Apcera Platform version 506.

  • Removed --hide-docker flag from apc package list (functionality previously removed).

  • Fixed bug in apc job logs which caused extra newlines to be printed between lines.

  • Fixed bug in apc app update where a scheduling tag could not be deleted on a job if it was the only scheduling tag present.

Web Console (1.20.4)

  • Fixed error messaging so that syntax errors in policy generate errors without loss of data.

  • Fixed bug so that removing scheduling tags now works as intended.

Orchestrator (0.3.3)

  • No changes.

Guide (1.0.72)

  • Updated these release notes.

  • Added release notes for this release.

  • Updated documentation for Health Manager heartbeats.

  • Documented how to use the group.allow claim for configuring LDAP groups.

Feb 08, 2016

Cluster (504.1.8)

  • Upgrade Notes
    • The release of Apcera version 504 contains many new features and fixes, and a critical kernel security update.

    • Apcera recommends that you upgrade to 504 as soon as possible during a maintenance window of your choice.

    • Note that 504 is a forward-only release. Once you upgrade to 504, you cannot rollback to a previous release.

    • To upgrade to 504, you must use Orchestrator version 0.3.3 which includes a critial security update.

    • After deploying 504, to complete the security update, you must update the Orchestrator host OS and restart each server in the cluster, one at a time.

    • Follow the Upgrading to Apcera release 504 instructions to upgrade to Apcera version 504.

  • Cluster changes
    • Fixed bug related to not being able to remove job scheduling tag.

    • Fixed bug related to fresh installation of 504.1.6 with Zabbix configuration failing.

    • Implemented requirement that release 504 should require 449.

    • Fixed bug preventing the automatic scaling down of Docker-based jobs.

    • Fixed bug that deleted policy document when there is a syntax error in the document.

    • The following features were released in version 504.1.6 (February 1, 2016).

    • Added support for pulling docker images during apc manifest deploy.

    • Added ability to authorize HTTP API requests via a URL query parameter.

    • Fixed bug with IP manager that prevented it from working reliably in clusters deployed across multiple subnets.

    • Fix bug in health manager causing it to ignore a job update and continue to enforce a previous state.

    • Changed all unmount system calls to use MNT_DETACH instead of MNT_FORCE (which can lead to data loss for NFS mounts).

    • Added support for http.client_acl and http.routing_tags config settings to API, auth servers (common/component).

    • Added colors and streaming events fixes for docker run command.

    • Fixed bug resulting in inability to authenticate via LDAP if no group memberships are present but the user is authorized via policy statements.

    • Corrected an issue so that jobs can now exceed their netmin value and burst up to the configured netmax value.

    • Updated apc changelog to now provide a link to the docs site's changelog (this page).

    • Fixed discrepancy between policy files and apc policy on command. The apc policy on command will display consistently with policy files after those policy files are updated.

    • Added --instances flag to apc app from package.

    • Fixed APC input parsing to correct issues with job updates.

    • Added claim type defaultRouteSuffix, which is used as the suffix in default route creation.

    • The default job route scheme no longer includes namespace, but does include app name, a short random string, and the domain.

    • On job policy check, update permission is required for all jobs that are using the route.

    • Miscellaneous fixes for Virtual Networking.

    • Fixed unhealthy target detection no longer removes the failing target from rotation.

    • When utilizing the network parameter on job create, permit join is required for job and network.

    • Fixed an issue where scheduling tags derived from policy would persist on a job after relevant policy was removed.

    • Refactored the 'SchedulingTags' field on job object to be an array of structs.

    • Enhanced dynamic generation of default namespace by adding defaultNamespacePrefix claim to auth resource and scrubbing of user name NS by replacing each unsupported punctuation character with an underline.

    • Updated apc cluster instance-manager list to right align resource columns.

    • Updated apc update behavior to not automatically output the changelog.

    • Added sorting of instances by their UUID on apc job instances / apc app instances.

    • Fixed an issue around updating metadata on Docker-sourced Packages.

    • Fixed an issue where sensitive information could be emitted in the instance manager logs.

    • Reduced some overhead in calculating component stats, generated every 5 seconds.

    • Added LDAP group compression.

    • Updated API server to prevent setting the num_instances of a job to 0, even when stopped.

    • Fixed an issue around policy scoping for Docker volumes.

    • Added apc manifest deploy command.

    • Fixed bug with APC websocket upgrade request, and enabled websocket origin checks in api server.

    • Remove continuum-setup source code and references thereof.

    • Fixed issue with apc app from package not applying the environment settings to the job.

    • Fix an issue with email attribute being retrieved inconsistently depending on the configuration.

    • Fix bug with policy validator regarding wildcards.

    • Default resources for Service Gateways has been increased. Each SG will now consume 16MB RAM and set the network cap at 100Mbps.

    • Fixed bug with stagpipe not validating a blank local name, and JM not using the BasicValidation method.

  • Chef changes
    • Remove hardcoded download from AWS of linux-image-extras.

    • Changed the kernel to 3.13.0-29.53.apcera4 version

    • Added a backup script for the NFS server role, which syncs a copy of the /data directory into S3 if a backup bucket is configured in the cluster.conf.

    • Shipped the zabbixapi gem file with chef-zabbix. This prevents the downloading zabbixapi from internet in air-gap installation mode.

    • Added the basic-auth-server component as part of the flex-auth architecture.

    • The Zabbix monitoring system will now be fully populated after the first deploy run on a new cluster or when adding monitoring to an existing cluster.

    • Upgrade all hosts in the cluster to the latest Linux kernel to address CVE-2016-0728. Reboots to all hosts are required.

    • Set TMPDIR to appropriate locations during component startup. This prevents the PM from placing temporary files in the root filesystem.

    • Fixed an issue where the orchestrator-agent process started during bootstrap configuration conflicts with the runit managed copy configured by Chef.

    • Purge the old continuum_audit database from the component-database server, which has been replaced by the dedicated auditlog database servers.

    • Enabled Package Manager routines for cleaning up the local disk usage.

    • Adjusted SCSI timeout configuration for better behavior in VMware environments.

    • Modified the syslog-ng configuration to correctly handle kernel timestamps.

    • Enabled MongoDB, S3, and rabbitmq service gateways in the default set of SGs configured by Stagehand.

    • Added an execution of the component-database backup script during deploy runs, to ensure a recent backup exists.

    • Enabled postgres database backups to local disk by default. Existing configuration hooks to store the backup in S3 still enable off-host backups.

Apcera Setup (1.3.2)

  • Apcera Setup 1.3 is compatable with Apcera version 504; earlier versions of Apcera Setup are not compatible with Apcera 504.

  • If you have an existing Apcera CE deployment, you should upgrade to Apcera 504 using Apcera Setup 1.3.

  • Made an enhancement to not duplicate IPs for multiple machine types.

  • Fixed process to run ntpdate-debian before deploy kicks off to make sure time is synchronized.

  • Improved reliability for SSH.

  • The following features were released in version 1.3.1 (February 1, 2016).

  • Added support for Orchestrator 0.3.3, and added reminder to restart the VMs due to kernel upgrade.

  • Added warning at deploy time to restart if images are old.

  • Optimized the certificate checking when using HTTPS.

  • Pinned stats collection to IM 1.

  • AWS installation now requires security group to allow port 443.

  • Added -r and -s flags to "install" command flags as in "deploy" command.

  • Added auto provision of NFS provider.

  • Fixed Windows file path handling.

  • Fixed "destroy" command clean up the config file even when the command fails leading to unrecoverable VMs. With -f flag you can have the original behavior.

  • Fixed issue where logs were truncated when using the "install" command.

  • The following features were released in version 1.2.3 (January 8, 2016).

  • Added support for custom DNS server to facilitate air-gapped installs.

  • Changed HTTPS certificate validation check to only prompt three times.

  • Changed Apcera-setup install to save config.json on failure.

  • AWS installation now requires security group to allow port 443.

  • Fixed error code returned on failure.

  • Updated the wording of the apcera-setup tool display messages.

  • Added support for user provided domain name.

  • Added support for user provided HTTPS certificate.

  • Added ability to automatically generate a self-signed HTTPS certificate for Apcera Platform internal use.

  • Added support for distributed switches in vSphere.

  • Re-organized the sectioning of the prompt and output.

  • Fixed issue using vSphere with a password that contains the pound sign character.

  • Fixed bug in concurrent create code, first two VMs are created sequentially now.

  • Changed to only create five VMs concurrently at a time instead of unbounded.

  • Changed to handle orphaned VMs in the vSphere inventory gracefully.

  • Fixed the destroy command Ctrl-C behavior.

  • Added support to creating VMs in parallel.

  • Added support for passing OpenStack Credentials through env vars.

  • Added filter for VirtualBox NICs to only those with IP addresses.

  • Fixed the OpenStack job creation issue.

  • Fixed apcera-setup destroy fail validation if suspended.

  • Fixed showing console URL before deployment.

  • Fixed state transition bugs.

  • Fixed error reporting for OpenStack error root causes.

  • Added vSphere support for creating a VM on a different host from the one on which the template was cloned.

  • Added vSphere support for providing a list of valid datastores to a VM.

  • Added support for reusing the provider config when the "config" command is run again.

  • Added support for OpenStack.

  • Fixed the "logs" command, it was broken on the previous release.

  • For AWS, fixed prompt message to ask for a VPC number instead of subnet.

APC (0.28.0)

  • You must use APC version 0.28 with Apcera Platform version 504 (you are prompted to upgrade when you target a 504 cluster.)

  • Updated apc cluster instance-manager list to right align resource columns.

  • Updated apc update behavior to not automatically output the changelog.

  • Added sorting of instances by their UUID on apc job instances / apc app instances.

  • Added apc manifest deploy command.

  • Fixed bug where apc docker run --no-start was working, but apc docker create was not.

  • Fixed issue with apc app from package not applying the environment settings to the job.

  • Updated APC package import behavior to remove failed packages.

  • Update APC to provide better help messages if no matching subcommand is found.

  • Fixed issue where apps created via apc app from package were not showing their start command as expected.

Web Console (1.20.0)

  • Created and delete providers from list page.

  • Made expected instances editable on status card.

  • Added provider description to provider list page.

  • Changed restart conditions from job detail view.

  • Added job subdivisions.

  • Added editable package envvars.

  • Added editable job labels.

  • Added gateway detail page.

  • Implemented gateway list page/Promote job to gateway.

  • Addec additional job metadata to runtime tab.

  • Added job label display.

  • Added combo box to include new item in freeform input.

  • Hid docker tab on packages list if empty.

  • Added network description to network detail page.

  • Changed runtime card title to 'Info'.

  • Don't suggest invalid namespaces in audit page filter select.

  • Updated to use consistent messaging for unsupported browsers.

  • Fixed mithril version.

  • Exposed table rows to consumers of job-table component.

  • Attached modal dialog event listeners only when open.

  • Added FQN to package detail.

  • Improved filter select UX on audit page.

  • Fixed policy upload bug.

  • Fixed to properly dispose ComboBox view event handlers.

  • Removed potential for empty selectors from /helpers.

  • Fixed use of empty selector in feature flag component.

  • Fixed uncaught mithril rendering error.

  • Reset saving state when user declines update.

  • Added show policy doc name validation message.

  • Fixed error when filtering namespaces.

  • Removed N+1 query when drawing namespace navigator.

  • Fixed issue to prevent false positives in namespace filter.

  • Fixed polling for new audit records.

  • Fixed job updates.

  • Updated timezone parsing regex to support IE.

  • Send labels as part of job json.

  • Fixed CookieStore and spec.

  • Added m.route to config attribute of stager link.

  • Fixed filter handler bug on /apps, and /capsules.

  • Fixed namespace filter suggestions overflowing menu

  • Don't render hidden notice.

  • Fixed serialization of routes on job.

  • Added middleware for converting JSON responses to model class.

  • Updated selenium web driver version to remove dependency issue.

  • Introduced option to define alternate key name to use when converting Form field to JSON.

  • Fixed test symlink dependency.

  • Rewrote kv list to use flexbox.

  • Don't store job resources as m.prop.

  • Refactored scheduling tag logic to reflect API changes.

  • Updated code to now ignore apcera-platform config file.

  • Upated code to use standard helper interface for icon helper.

  • Made InlineInput component improvements.

  • Revived javascript browser testing framework.

  • Removed process name from runtime info

  • Updated spacing between sidebar keyline

  • Updated keymaster to ease testing

  • Refactored related policy component interface

  • Updated to use Agency on policy pages.

  • Made inline input into component.

  • Standardized format of dates.

Orchestrator (0.3.3)

  • Forced upgrade. You must upgrade to Orchestrator 0.3.3 to deploy Apcera version 504.

  • Version 0.3.3 includes a package that upgrades the Linux kernel for each Apcera server with critical security fixes.

  • After deploying 504, to complete the security update, you must update the Orchestrator host OS and restart each server in the cluster, one at a time, during a maintenance window of your choice.

  • Follow the Upgrading to Apcera release 504 instructions to upgrade to Apcera version 504.

  • Added mechanism to warn user with a release specific message (optionally defined by the release bundle), and let user decide to proceed or not.

  • Added flag to run orchestrator-cli in non-interactive mode which proceeds with the installation without attending to the warning.

  • Added 'allowed AMIs' list to enable patching existing instances instead of being forced to upgrade when a new AMI is specified in release.json.

  • Orchestrator version 0.3.2 was released internally on January 6, 2015 with the following features (available in 0.3.3).

  • Changed ordering of how vSphere definition returns the IP address.

  • Changed the vSphere machine definition to prioritize returning the chosen IP address from its list of networks rather than returning it from the vSphere guest tooling.

  • Changed branding from Continuum to Apcera.

  • Added support for passing multiple cluster.conf files using the -c command (or --config) to deploy a single cluster, with the last file taking precedence.

  • Fixed potential space overrun issue; you must now run Orchestrator as the Orchestrator user.

  • Added logic to better handle invalid SSH selection.

Guide (1.0.71)

Dec 04, 2015

Cluster (449.3.1)

  • Cluster changes
    • Fixed an issue where stagehand would apply job updates intended for a single stager to multiple stagers

    • Added group whitelist and escaping in policy.

    • Support v2 Registry API of the public Docker Hub.

    • To access the public Docker Hub, the docker.allow policy must now use the HTTPS URL // instead of //

    • Deprecate consumption of the v1 Registry API of the public Docker Hub.

    • Quay, Artefactory, and other v1-compatible registries are unaffected by this change.

    • Fixed an issue handling of job restart modes occassionally causing extra instances to be launched.

    • Fixed an issue with scaling operations on a job with a semantic pipeline incorrectly causing the existing instances to be cycled.

    • Remove stager-based method of Docker image creation.

    • Remove docker.lockPublicBaseImages policy claim and handling.

    • Remove docker.cache policy claim and handling.

    • Docker image layers created using the new method will be cached securely by default.

    • Remove GET /preferences/job API handler.

    • Remove POST /jobs/docker-job-check API handler.

    • Add / and * to whitelisted characters which can be part of a route.

    • Added join for permit all policy on job.

    • Fixed container initd process to respect all user's group memberships when starting a container for that user.

Apcera Setup (1.1.2)

  • Added support for filtering out incompatible hosts from a vSphere cluster when picking a random host.

  • Added support for vSphere version 5.5 for provisioning cluster environments.

  • Removed limitation on the number of Instance Managers during apcera-setup configuration.

  • Added support for running apcera-setup on Linux.

  • Resolved issue with "not authorized" user on VMware.

  • Miscellanous bug fixes.

APC (0.24.0)

  • Fixed the apc download flow on Windows and also download the zip file for apc.

  • Changed apc docker run to use the improved Docker image creation by default.

Web Console (1.19.0)

  • Add unsupported browser messaging

  • Add job delete to job list

  • Remove Job from route

  • Include all event types in suggestion set on audit log

  • Add line count toggle to tail log feature

  • Remove unused inputs for adding job to network

  • Create / Delete Network

  • Delete service from service details

  • Delete network from network details

  • Add default column widths for all tables

  • Miscellaneous UI cleanup and polishing.

  • Fix audit log time filter

  • Update service binding to use new API endpoint

  • Disable browser autocomplete for all form inputs

  • Fix provider detail redirect

  • Fix keyboard handling with inline edit component

  • Update package list after package delete

  • Fix cookie support library for IE9

  • Fix namespace nav counts

  • Reload policy

  • Fix inline edit bug

  • Check open port list when enabling ssh (fixes bug with opening port twice on capsule)

  • Fetch resource after updating in all cases

  • Add confirmation to policy doc delete

  • Inline logo svg (workaround to server bug)

  • Fix metrics for clone jobs

  • Redirect to list view in case of API 404

  • Fix filter bug on audit page when start/end times blank

  • Fix page level notice in ie10+

  • Fix select box in IE10+

Orchestrator (0.3.1)

  • No changes.

Guide (1.0.55)

  • Updated Docker documentation for supporting Docker Hub API 2.0

  • Removed deprecated Docker job creation functionality

  • Updated policy seal examples

  • Updated policy input claim types and values table

  • Updated apcera-setup v1.0 documentation

Nov 23, 2015

Cluster (447.2.0)

  • Cluster changes
    • Note that the system policy clusterPermissions.pol is missing the claim permit join for the job resource. Please manually add this claim to this policy for this release.

    • Added --depends-on, --pkg-env, and --copy flags to apc app deploy.

    • Added --client-acl to apc app create/update, apc app from package, apc capsule create, apc docker run, apc job update.

    • Removed deprecated concept of the job weight.

    • Removed deprecated NATS subjects and apcera.cntm.router.update.job.*.

    • Fixed documentation on apc pipe commands.

    • Fixed documentation on --client-acl flag in job creation and update commands.

    • Fixed API server handling of pre-existing, invalid job routes.

    • Fixed a bug in APC filecopy causing potential file corruption.

    • Added --new-api flag to apc docker run for newer, more performant, server-driven approach. Currently in beta.

    • Added --docker-image flag to apc package list to show packages satisfying a Docker image, produced via the --new-api approach.

    • Added apc network commands to create and manage virtual networks for jobs.

    • Added --compliance flag to app [job, app, gateway, stager] show to indicate whether the appropriate object is policy compliant.

    • Added ability to list jobs by labels to apc job list.

    • Changed route weight displays in apc.

    • Changed how to SSH into a job to the reliably load shell environment.

    • Fixed handling and documentation for the --netmin flag (-n) on misc apc commands.

APC (0.23.0)

  • Added logging to websocket requests.

  • Added documentation for job show compliance flag.

  • Added command to display route weights.

  • Added --new-api flag to trigger new Docker API flow (for example, apc docker run -i <image-name> --new-api myJob).

  • Use common helper for job deletion.

  • Added warning that Windows APC does not support Cygwin.

  • Added support for -it and -rm flags.

  • APC Netmin flag short version.

Web Console (1.18.0)

  • Added network support.

  • Added provider support.

  • Separated namespace nav and jobs table filters.

  • Improved job and instance level resource quota/utilization stats.

  • Added affinity to IM scheduling tag form.

  • Use new and correct API endpoint for service show

  • Added 'data center' IM attribute where IM's are displayed.

  • Added suggestion list to bind to service form.

  • Added confirmation gate to delete actions.

  • Added create service option.

  • Ensure live update on all api resources (Agency polling).

  • Use shared cache (Agency) for Packages, Services, Job details.

  • Rebranded with new logo

  • Added select an open port option to job link form.

  • Added suggestion list to job link target.

  • Added filtering to routes list.

  • Organized Job Details content by tabs.

  • Made visual and minor updates.

  • Improved browser support for IE10 and FireFox.

  • Fixed bug to properly escape combo box input

  • Fixed bug to properly parse certain api error responses.

  • Fixed bug to fully hide suggestion list when empty.

  • Fixed bug to respect start time on audit log time range input.

  • Fixed bug to link to correct favicon url from login dialogue.

  • Fixed bug to stop polling when leaving staging page.

Orchestrator (0.3.1)

  • Converted to syslog-ng and setup syslog target to feed logs to Splunk.

  • Updated Orchestrator DB.

  • Removed virtual memory limits for performance benefits.

  • Added public IP addresses to availability list for non-AWS deployments.

Guide (1.0.54)

  • Updated list of policy input and output claims per resource type.

  • Updated fonts/colors to match corporate site and rebranding.

  • Added note to App Virtual Networks that UDP is supported.

  • Updated Orchestrator documentation.

  • Updated blue-green tutorial.

  • Enabled full-text search (the Chrome browser is recommended for speed of search).

  • Fixed bugs in apcera-setup docs.

  • Added Orchestrator DB backup and recovery procedure.

  • Telling user to upgrade VMware vms if prompted (changed to don't upgrade).

  • Added Max number of IMs to setup home page.

  • Fixed the title for the docs from Apcera Hybrid Cloud OS to Apcera Documentation.

  • Added explanation to RNs regarding change to default Crowd against email (ZD1204).

  • Updated the Hybrid cloud example.

  • Add Job Label parameters to Job API endpoint documentation.

  • Added Oracle Service Gateway documentation.

  • Added note to Service Gateway documentation about the stateless version being deprecated.

  • Fixed LDAP Auth policy example.

  • Added note to S3 gateway regarding bucket deletion.

  • Updated Virtual Network documentation.

  • Updated nav titles to better distinguish between Apcera Platform setup and production deploys.

  • Published Docker API documentation.

  • Added example on how to use templates at job creation time.

  • Updated docs with new Apcera logo and name.

  • Fixed miscellaneous typos, spellings, keywords, links, etc.

  • Version 1.0.53 follows.

  • Added Apcera Platform Setup documentation.

  • Added Service Gateway API documentation for stateless and async.

  • Added RabbitMQ SG documentation.

  • Added filecopy and scp content.

  • Updated Orchestrator-cli update procedure.

  • Updated Walkthough tutorial and APC install docs.

  • Added APC log in steps to Install and Use APC.

  • Created Hybrid Cloud Simulation tutorial.

  • Moved Policy API documentation to API endpoint list.

  • Fixed miscellaneous typos, spellings, keywords, links, etc.

Nov 03, 2015

Cluster (444.2.7)

  • Cluster changes
    • Deprecated stateful Service Gateway API (replace with stateless).

    • Added new Service Gateways for MongoDB and S3.

    • Updated Service Gateway API to support stateless and async operations.

    • Added Datacenter field to InstanceManagerStats API.

    • Add support for groupofNames apart from POSIX groups in LDAP.

    • Implemented static policy validation for labels.

    • instance_manager dedup resources to download by digest.

    • Implement job list filter-by-label.

    • api server added service by uuid endpoint.

    • Replaced admin with search for LDAP config.

    • api_server new multi resource upload download API.

    • package_manager - only remove resources no longer in use.

    • Misc code cleanup.

APC (0.22.0)

  • Fixed nil deference panic in capsule create.

  • Rollback APC version filter to not return the incompatibility error.

  • Enable new fields for apc to display when dealing with async gateways.

  • Enhanced apc login.

  • Fixed APC Job Health Checks.

  • Remove outdated apc s3 urls.

Web Console (1.17.0)

  • Add evaluation agreement requirement for certain clusters.

  • Use key value list cards instead of multiple key value cards for details pages.

  • Display job tags as attributes.

  • Link to apc binary's on cluster.

  • Improve API error message display.

  • Standardize route metric display.

  • Add default sort order to all lists.

  • Update current filter styles.

  • Add job status to job details, job list, route details.

  • Display route weight percentage on route details.

  • Add effective policy page - displays policy by realm/namespace.

  • Add jobs using this package table to package details.

  • Add resizeable headers to all tables.

  • Remove start stop from stager details.

  • Fix cache max size.

  • Standardize font color across certain pages.

  • Disable autocomplete on combo box.

  • IE11 graph clipping bug.

  • Fix bug with polling causing a spinner every 10 secs.

Orchestrator (0.3.1)

  • Add Proxy Protocol support to Router.

  • Set the cluster name var in api_server conf.

  • Pull in new version of chef-graphite.

  • Add mounts.rb entry for Zabbix Postgres.

  • Changed default policy for crowd auth to test against user 'name' instead of 'email' because crowd users can modify their email.

Guide (1.0.52)

  • Inlcudes 1.0.50 and 1.0.51.

  • Fixed broken links.

  • Changed demo to trial image.

  • Removed FAQs from nav.

  • Removed mongoDB from RNs for 442.

  • Misc minor fixes (typos).

  • Updated Swagger schema for Apcera API.

  • Added Docker API documentation for private build.

  • Added draft Hybrid cloud config documentation.

  • Added Apcera trial cluster documentation for beta.

  • Documented new service gateways for RabbitMQ, S3, MongoDB.

  • Updated Apcera API documentation.

  • Added API ref, model, recipes docs.

  • vSphere install update for using SSH.

  • Document Orchestrator version 0.3.0 (forced upgrade).

  • Added Job Labeling content.

  • Added using arbitrary strings for policy claims.

  • Added supported claims type table.

  • Added SSH and app console topic.

  • Updated vSphere requirements.

  • Added runtime template tutorial.

  • Updated Support Service Types table with usage details.

  • Updated manifest file example.

  • Miscellaneous minor bug fixes.

Oct 21, 2015

Cluster (442.8.3)

  • Cluster changes
    • APC is now distributed through the API server.

    • Added support for multiple cluster subnets and extended the network service to support multiple subnets via the ipnets paramter.

    • Added LDAP group membership support.

    • Added better error messages for NFS service gateway failures.

    • Fixed spelling errors in log messages.

APC (0.21.0)

  • Added ability to define labels for jobs, which can get set either by policy, or when calling "apc job update" with "–label-set" option.

  • Added options to delete services when deleting a job with "–delete-services" option.

  • Fixed issue with importing multiple packages on Windows

  • Fixed issue with properly calculating the uptime for running instances

  • Fixed an issue with retrieving job health before the information was available

  • Updates to allow network services to expose multiple network ranges

  • Improved handling of "apc update" to keep in sync with the proper version for the cluster.

  • Updated apc help for namespaces

Web Console (1.16.2)

  • Shared resource cache: im's, job list, job detail

  • Added 'state' to package list

  • Added restart button to job details

  • Added CRUD for ports on job details

  • Added ability to filter packages by namespace and state

  • Added namespace to jobs list filter suggestions (decoupled from namespace navigator)

  • Prompt user to restart job after changing ssh option

  • Improved UX on instance kill; job details, im details (spinner, disabled viz)

  • Improved API error message handling

  • Improved keyboard handling on filter input

  • Improved error handling on job 404

  • Added ability to kill an Instance from instance manager details

  • External link on route list; route list, route details

  • Improved error handling on internet disco (laptop sleep, e.g.)

  • Improved style on links in tables (jobs)

  • Improved loading state indicator (cluster)

  • Use name instead of fqn for service bindings on job details

  • Updated confirmation message on service bind delete

  • Improved client-side validation on job link

  • Added some transitions on form open

  • Requests/sec. and Errors/sec now rounded to 1 decimal.

  • Added abilility to choose start/end datetime on audit log component

  • Fixed issue with tab param routing

  • Fixed metrics query issues

  • Fixed package details 'stuck' issue

  • Standardized running jobs/instance between jobs/cluster

  • Fixed issue adding a first route to a job

  • Prevent navigation when clicking an active tab

  • Fixed issue flash of login page on page refresh (erroneous 401)

  • Fixed issue with empty service list in service binding form on job details

Orchestrator (0.3.0)

  • Orchestrator version 0.3.0 is a forced upgrade. Refer to the Orchestrator section of the docs.

  • Fixed issue with properly validating checksums of release assets.

  • Added support for minimum validation being specified in releases.

  • Updated releases to be retrieved from new default location.

  • Added ability to collect logs from specific locations.

  • Softlayer provisioner has been renamed to generic.

  • Improved error reporting when deploy is ran before the init command.

  • Fixed an issue where EBS volumes could be attempted to be attached to multiple hosts.

  • Increased timeout orchestrator waits after a host installs a new orchestrator agent.

  • Fixed issue with machine scale down where the machine metadata may be removed before the Chef teardown actions are completed.

Guide (1.0.49)

  • Product architecture overview

  • App Virtual Networking (DRAFT)

  • MongoDB Service (new service gate type)

  • LDAP/S Service Using Generic SG

  • Reorganized and updated install docs to include separate top-level sections for deployment, config, install, ops

  • Video reference library

  • Updated supported environments and software

  • App selection and migration

  • Instrumenting Java apps with New Relic

Sep 28, 2015

Cluster (440b)

  • Cluster changes
    • Instance Manager - Refactor DNS server to allow a-record-callbacks.

    • Package Manager - remove upload retry behavior.

    • Add better Package manager LRU cache.

    • Added Task Events framework.

    • APC app from package does not respect user's input. ENGT-3998

    • Add Digest support to packages.

    • Provide version and URL when updating apc. ENGT-3851

    • Rename the old package with timestamp. ENGT-3639 (resolves support ticket 827)

    • Check for service in use using Mfqn vs asking the gateway.

    • Policy events for audit log.

    • Remove scheduling tags from packages.

    • Split HM "starting timeout" into two separate timeouts.

    • Better error message for hard scheduling constraint. ENGT-3655

    • Job Link names must conform to ENV variable name constraints. ENGT-3703 (resolves support ticket 819)

    • Enhance evaluation of app detection rules from the api server.

    • Fix package list of an app after multiple app deploy.

    • Provide new default permission for all users to read cluster properties. ENGT-3288

APC (0.20.0)

  • Use SHA256 instead of SHA1 internally for package resource verification.

  • By default we delete old packages during an app deploy unless –keep-previous is specified.

  • Windows apc no longer requires a dependency on cygwin.

  • Various bug and UI fixes. (resolves support tickets 768,773,807)

Web Console (1.15.0)

  • Add audit tab to policy doc list.

  • Add more explicit copy to job link & job route forms.

  • Link to route detail from job detail routes table.

  • Various style changes.

  • Fix flash of content on load more button on audit page.

  • Hide job links for service gateways.

  • Fix namespace navigator combo box select.

  • Add key/value table component.

  • Add gradient to primary action buttons.

  • Adjust color scheme of side nav.

  • Add html title tag to more cell types in table component.

  • Improve loading state indicators, and apply to more places.

  • Add toggle for allow_ssh attribute of job.

  • Make service description sort-able.

  • Move docker packages table to their own tab.

  • Prompt user to restart job on selective job updates.

  • Add inline edit component.

  • Move instance attribute to k/v card and make editable on job details.

  • Persist filters when navigating back to list from details.

  • Add create/delete job link feature.

  • Add job attract/repel feature with job suggestions.

  • Add combo box component.

  • Allow arbitrary filters in filter select.

  • Improve error handling on cluster page

  • Fix linking to sorted table

  • Fix policy doc upload

  • Fix suggestions in namespace navigator

  • Fix various issues with route metrics

  • Remove cache of old resources when navigating

  • Correctly escape user input on filter select

  • Fix link to policy doc from policy in tabs

Orchestrator (0.2.3)

  • Updated sample configuration files.

  • Updated Orchestrator CLI help.

  • Improved cluster teardown operations.

  • Updated error message if init is not run.

  • Added auditlog component to Terraform config.

  • Added support for auditlog-database role hosts to have a separate disk mounted at /var/lib/postgres. This must be set up at provision time.

  • Note that you must deploy the new auditlog-database component in your cluster, or the new web UI will not show audit logs.

Guide (1.0.49)

  • Misc bug fixes and improvements

  • Added examples for Zabbix email alerts

  • Added example for air-gapped package management

  • Added example for NFS provider registration

  • Added example to use apc service show to display network details (ENGT-3013)

  • Added examples for Zabbix email alerts.

  • Added example for air-gapped package management.

  • Added example for NFS provider registration.

  • Added example to use apc service show to display network details (ENGT-3013).

  • Release v1.0.48 that includes the following items.

  • Added config examples to Monitoring Your Cluster.

  • Updated Google Oauth documentation.

  • Added LDAP groups documentation.

  • Updated Terraform installation documentation.

  • Updated OpenStack installation documentation

  • Fixed issue causing documentation to load slowly (disabled full-text search; re-enabled keyword search only).

  • Added exposed/reserved ports documentation.

  • Reorganized the deployment/installation section of the documentation.

  • Updated Orchestrator usage examples, including orchestrator-cli update command.

Sep 16, 2015

Cluster (438g)

  • Cluster changes
    • Split package cache optimization (ENGT-742).

    • Enable cleanup on delete for local cache (ENGT-1072).

    • Inverted docker.allow default to deny all without any docker.allow policy in a namespace.; added "*" wildcard to allow all; added commented policy to the defaultPolicies directory provide examples for a user to easily get started running Docker images (ENGT-2493).

    • Allowservice show to display parameters whitelisted in policy per-gateway (ENGT-3013).

    • Added the ability to add hard scheduling tags to the staging coordinator through the use of policy on a namespace. Also removed the previously added code which tagged stagers with the job uuid of the staging coordinator (ENGT-3111).

    • Fix error during SendGroupNotify (ENGT-3411).

    • instance_manager – Use async sleep in start_help_request (ENGT-3587).

    • Adjust timings in instance manager for even instance distribution (ENGT-3589).

    • Add a check after the dynamic binding timeout to ensure we still care (ENGT-3644).

    • Jobs which are stopped will no longer get restarted during a run of stagehand (ENGT-3648).

    • Clarify that binding has succeeded, but the job failed to start (ENGT-3701).

    • Fixed API server crash on inconsistent data (ENGT-3760).

    • Added a stateful 2-tier service gateway for MongoDB (ENGT-3766). Disabled for now.

    • Fixed instance manager metrics when no containers are running (ENGT-3778).

    • Added ngx_http_real_ip module to the nginx build (SRE-232).

    • Code cleanup – removed RequestNoError.

    • Code cleanup – instance_manager (dynamic) bindings cleanup and commenting.

    • Code cleanup – added logging of request_id on all API server calls.

    • Code cleanup – added logging of the instance UUID to dynamic bindings.

    • Code cleanup – Added the instance's host to the HM's heartbeat log line.

    • Reverted to use the slightly older GNATSD INFO message format.

    • Bump timeout on EnsureAllInstancesUpdated from 10 to 30 seconds when connecting to a capsule.

APC (0.19.6)

  • Refactor how .apc files are written to prevent corruption (ENGT-69, ENGT-3445).

  • HTTP scheme is no longer the default in add/delete/update route commands. Users must explicitly state whether to use an HTTP or TCP route, either by writing it in the endpoint field or by using the explicit boolean flags. This makes it much easier to describe intent, that is, '–http' is no longer implicitly true (ENGT-3590).

  • Improved error message when we add/delete/update a route on a job and the port it is intending to serve is not exposed (ENGT-3108).

  • Windows uses %USERPROFILE% instead of $HOME (ENGT-3423).

  • Fix docker help message (ENGT-2475).

  • Improved route type auto-detection.

Web Console (1.13.0)

  • Add service details page

  • Add metrics to route table (namespaceStats feature flag)

  • Update to Service Binding create on job details

  • Updates to running / total jobs counts on ns and cluster pages

  • Link to package details from job#packages card

  • Improve visual style of graph cards

  • Show more error messages on job details

  • New audit pagination (newAudit feature flag)

  • Add skanderna as acceptance cluster support

  • Loading state to modal dialog component

  • Delete item button visual improvements

  • Weekly visual change tweaks (design team bucket list)

  • New audit log visual improvements (newAudit FF)

  • Temporarily disable policy validation (preview) feature (backend bugs)

  • Add sticky table headers

  • Add route create to job details

  • Show loading spinner instead of blocking page on api requests by default (background)

  • Enable Kerberos login

  • Feature flag namespace health feature.

  • Fix dismiss handler on dropdown component

  • Fix memory leak on cluster details page

  • Visual style for anchor links

  • Logout user out on any 401

  • esc key closes dropdown

  • Fix various query issues with namespace stats (feature flagged)

  • Link to jobs table from stats cards (namespaceStats feature flag)

  • Fix single form based auth login bug

Orchestrator (0.2.3)

  • Same as previous release.

Guide (1.0.47)

  • Updated Docker whitelist documentation. Docker is entirely controlled by policy. To run or create a Docker job, you must create policy.

  • Updated NFS documentation to support new NFS implementation.

  • Updated Supported languages and frameworks section.

  • Fixed Miscellaneous typos

Sep 09, 2015

Cluster (437h)

  • Cluster changes
    • Temporarily removes the INFO message format changes from gnatsd to enable deployments over the older version of nginx-router/nginx-nats.

    • Fixed repositories.conf to pick up the patched gnatsd

APC (0.19.5)

  • Same as previous release.

Web Console (1.11.1)

  • Same as previous release.

Orchestrator (0.2.3)

  • Same as previous release.

Guide (1.0.46)

  • Added local NTP server config example for vSphere deployments.

  • Added description of the manifest file format.

  • Updated ActiveDirectory documentation to use generic cluster name.

  • Updated Google Oauth documentation to better indicate where to enter secrets.

  • Published new Apcera API documentation.

  • Fixed typo in LDAP documentation.

  • Fixed typos in Policy Permissions documentation.

  • Added link to providers documentation in the Walkthrough tutorial.

Sep 02, 2015

Cluster (436h)

  • Cluster changes
    • Benchmark Current Implementation of Volume Manager (ENGT-2610).

    • Fixed bug with cluster not keeping track of available resources properly (ENGT-2748).

    • Improved package creation failure messaging (ENGT-2864).

    • Incorrect handling of Yottabyte postfix in bytesize parsing (typo in code) (ENGT-2890).

    • Enabled multiple datacenter support for the HTTP router (ENGT-2966).

    • Bindings included with job from jobs/:uuid endpoint now include target job name (ENGT-3130).

    • The 'health' param now exists on the '/jobs/:uuid' endpoint (ENGT-3142).

    • Added "description" metadata field to policy documents (ENGT-3143).

    • Added error checking for job health: nats timeout (ENGT-3419).

APC (0.19.5)

  • Improved the APC error message for command not found (ENGT-1784).

  • Fixed the issue that causes APC to fail with LDAP authentication when there is no .apc file (ENGT-3435).

  • Document what "unpredictable behavior" means for job linking to multiple instances (ENGT-3171).

  • Command 'apc job start' displays the status of how many have started, e.g., 45/1000 instances running (ENGT-3209).

  • APC version 0.19.4 was a hotfix for a segfault while reading passwords on Windows.

Web Console (1.11.1)

  • Change login and help buttons to primary action

  • Skip policy preview to work around job link deletion bug

  • Move persistence logic from api model to view model

  • Fix sorting by uptime on job list on IM detail page

  • Always stop polling when user calls resource.poll(false)

  • Allow deleting running job with confirmation

  • Redraw after changing state successfully

  • Use s/s as unit for CPU

  • Properly reset instance count on cancel

  • Allow sorting by updated time on policy list

  • Iterate on Form Field Design

  • Add service binding add to job details

  • Add page level notification component

  • Add feature flag

  • Add remove binding to job details

  • Removes 32 bit apc binary download link

  • Update button styles

  • Include full table cell contents as tooltip

  • Provide external link for routes in routes table

  • Show running/all job counts on job list/state

  • Add add scheduling tag to job details

  • Add notification for package/job delete

  • Add modal confirmation component

  • Require confirmation for destructive actions

  • Add remove scheduling tag to job details

  • Link to polDoc in view mode instead of edit

  • Update metrics query

  • Add basic route list page

  • Add basic route details page

  • Add search to namespace nav on audit page

  • Round req/err metrics to one decimal

  • Display policy doc messages in notice component

  • Use shared cache for API resources (resolves double requests)

  • Right align numerical columns on tables

  • Fixed some job detail metrics

  • Fixed non json error response handling

  • Fixed several filter related bugs (ns and table filters)

  • Fixed job count on cluster page

  • Fixed job update state transition

  • Fixed namespace health indicator

  • Fixed route metrics on jobList/stats

  • Fixed graph time axis

  • Fixed filtering new audit log entries

  • Removed network reservation from graphs

  • Fixed LoginPopup grant type bug when only one selectable auth type is available

  • Fixed url encoding issue with multiple '.'s

  • Fixed duplicate name filter

  • Fixed binding issue on services list page

Orchestrator (0.2.3)

  • Orchestrator version 0.2.3 is now publicly available. Please update to the latest version.

  • See previous entries and Orchestrator documentation for release notes item. In addition, note the following additions that were previously not documented.

  • There is a new log collection facility for collecting component logs. new log collecting function. After upgrading to Orchestrator version 0.2.3, run command the orchestrator-cli collect logs all. This will log into each HCOS node and grab the necessary logs and place them into a tarball for each host on the Orchestrator. (Support ticket 644)

  • There is a length restriction for the base cluster domain to 62 characters (due to SSL certificate CN limitiation to 64 characters). (Support ticket 818)

Guide (1.0.45)

  • Made improvements for mobile viewing.

  • Updated Google Oauth documentation.

  • Added installation instructions using Terraform.

  • Added documentation for tagging Package Managers for hybrid cloud.

  • Added documentation for tagging stagers and staging coordinators for hybrid cloud.

  • Added policy example that demonstrates two-factor authentication.

  • Added documentation for applying the negation tab for hybrid cloud.

  • Updated search capabilities within the documentation.

  • Added documentation for vCloud installations.

  • Updated OpenStack installation documentation.

  • Miscellaneous updates to LDAP and Orchestrator documentation.

  • Added link to the Poliy Language Fundamentals policy.

  • Added note about IM disk accounting to the sizing guidelines.

  • Added documentation supporting Java JAR stager.

  • Added documentation explaining how to use Generic Service Gateway.

  • Added documentation supporting the use of the JDBC-specific env var.

  • Made miscellaneous improvements and bug fixes to the docs.

  • Added link to job scheduling video.

  • Added TCP examples for app routes.

Aug 19, 2015

Cluster (434b)

  • Cluster changes
    • Add pagination to API for audit.

    • Added metrics api performance improvements.

    • Updated staging job creation and management.

    • Added ability for 'apc docker run' to create routes upon creation.

APC (0.19.3)

  • Mac apc installer creates restricted directory.

  • App delete will look for manifest.

Web Console (1.10.0)

  • Add hover state to line graphs - including current value

  • Improvements to policy editing (disabled while saving,

  • Add job links readonly to job details

  • Add service bindings readonly to job details

  • Visual improvements to policy preview error states

  • Add readonly 'allow egress' as a flag (in addition to service binding)

  • Only show login options that are enabled in the cluster

  • Hide cluster tab if user doesn't have cluster permissions (causes a refresh issue - temporary)

  • Add 'delete job' feature

  • Add 'delete package' feature

  • Add 'kill instance' feature

  • Hide graph warning line when it overlaps with res/quota line

  • Add polling for job instances on job details page

  • Deep link tab navigation

  • Show scheduling tags in table on job details (readonly)

  • Hide sidebar nav after navigating when sidebar is hidden

  • Include y axis labels when data is all one value in graphs

  • Reset route on logout (to prevent permission issue when switching users)

  • Validate instance count in instance count input

Orchestrator (0.2.3)

  • Same as previous release.

  • Note that 0.2.3 is internal only. Version 0.2.1 is the latest promoted release of Orchestrator.

Guide (1.0.39)

  • Added note to Semantic Pipeline documentation

  • Added embedded video links

  • Updated Orchestrator version info

  • Added app from package for Postgres example

  • Added NewRelic app integration tutorial

Aug 13, 2015

Cluster (433c)

  • Cluster changes
    • Patch release.

    • Fix for /tmp permssions.

APC (0.19.2)

  • Same as previous release.

Web Console (1.9.0)

  • Same as previous release.

Orchestrator (0.2.3)

  • Same as previous release.

Guide (1.0.37)

  • Same as previous release.

Aug 04, 2015

Cluster (432j)

  • Cluster changes
    • Move job related policy to the Job Manager.

    • Quota and package policy checks skipped if job transitions to stopped state.

    • Default policy changed to accommodate stagehand and staging coordinator.

    • Search upstream servers when resolving AAAA records.

    • Add support for aggregation of job metrics by namespace.

    • Align metric timestamps on metrics interval.

    • Account for loop duration when sending metrics.

    • Allow semantic pipeline deletion in errored or finished state or if parent is gone.

    • Instance manager can now be configured to download packages from package managers inside the local datacenter.

    • Volume management extracted into separate component.

    • Corrected caching behavior of Docker layers in apc docker run.

    • Fixed file descriptor leak in the initd server. This fix requires job instances to be restarted to take effect. Long running processes may run out of resources and experience failures unless they are restarted.

    • Enabled descendent namespace querying for the new audit log (which is dark launched).

APC (0.19.2)

  • Fixed bug with deleting an errored job that has a semantic pipeline.

  • Corrected apc export all messages.

  • Introduced stager debugging for apc app create.

  • Added apc provider delete --force option.

Web Console (1.9.0)

  • Removes NS dashboard and stats from jobs table (pending metrics backend support)

  • Add color to stats graphs

  • Add gradient fill to stats graph

  • Improve display of graph legends

  • Persist UI state across pages/tabs

  • Job state now sorted by priority instead of lexicographically

  • Guard against bad data from metrics API (makes graphs less brittle)

  • Display Packages as a table on job details

  • Add 'policy in context' feature to most resources (this shows policy (including inherited policy) that applies to the resource or namespace you're looking at)

  • Dark launch new audit log (available at /audit/logs-new - must be enabled in cluster)

  • Display routes in a table on job details

  • Add cluster name to cluster page header

  • Make card title color consistent

  • Fix bug with URL encoding

  • Show resource reservation on job detail stats

Orchestrator (0.2.3)

  • Added new mechanism for adding datacenter tags to components based on CIDR ranges in the cluster.conf, and adding support for using those tags in the existing IM tagging, etc.

  • This also adds ability to override host DNS settings on based on CIDR or datacenter tag.

  • Use orchestrator.conf in current folder if one exists.

  • In the Conductor, prevent deploy from updating the release unless asked.

  • Better error detection if VM template is malformed.

  • Ignore the host already configured error if the key is not changed.

  • Display more useful error when servers shut down.

  • Added support for AWS EBS volume types "gp2" and "io1" with IOPS.

  • Ignore existing vNICs attached to continuum-template.

Guide (1.0.37)

  • Reorganized the core "Working with " topics into 4 categories (jobs, staging/packaging, services, and policy).

  • Separated concepts and usage/reference information from tutorials and how-to instructions.

  • Added lab exercises developed for the training and converted them into stand-alone tutorials, and added new tutorials.

  • Added documentation for installing the product in production.

  • Updated and reworked the policy documentation.

  • Added +/- navigational elements to the nav pane.

  • Created doc sites for enterprise docs.

Jul 20, 2015

Cluster (430c)

  • Cluster changes
    • Quay support

    • Make IM rlimits configurable

    • API server timeout to graphite increased

    • Docker job policy endpoint

    • Fix bug in disk accounting

    • Fix bug around container startup in the networking that could deadlock an Instance Manager

    • Have service name pull from FQN localname

APC (0.19.1)

  • Allow for private docker registries with basic auth

  • Automatically restart app if 'app update' fails

  • Improvements to help messaging

Web Console

  • Removes NS dashboard and stats from jobs table (pending metrics backend support)

  • Use grey background for header & sidebar

  • Update labels on dashboard page

  • Improve LineGraph "no-data" state

  • Add stats graphs to IM details page

  • Split job logs, graphs, and summary to tabs on job details

  • Removes policy-defined namespaces from NS nav (removes empty nodes)

  • Removes "30 Minutes" option from graph timelines (to help guard against data inconsistencies)

  • Use 'max' function to summarize graph data instead of 'avg' (to help sawtooth metrics bug)

Orchestrator (0.2.2)

  • Miscellaneous minor bug fixes

  • Added chef support for LDAP


  • Added support for using private Docker registry

  • Added support for LDAP authentication

  • Added support for Crowd authentication

  • Miscellaneous minor bug fixes

Jul 15, 2015

Cluster (429d)

  • Cluster changes
    • Metrics API now handles namespaces ending in a '/'

    • Auth server logs a warning if the default auth type is not configured

    • Instance manager does not try to launch containers when encountering disk failures (or /var/lib/continuum is read-only)

    • TBYB generated config now enables 'basic' auth

    • Metrics manager only shows total bytes in and out, not per interface

    • nginx-nats now handles SIGHUP without leaving zombie processes behind

    • Auth server UI now uses "Continuum Web Login" instead of "Apcera Web Login"

    • Instance manager allows nameservers with nonstandard ports

    • Added JDBC env variables for mysql and postgres bindings

    • Auth server: removed the default google web_client_id

    • Auth server verifies the LDAP configuration by connecting with admin creds

    • Instance manager: change spacing between sending metrics to the server

    • Metrics API keeps the timestamp even with a null value from graphite

    • Added stager debugging output

    • Support for defining optional ports through manifest

    • Upgraded openssl to 1.0.1p

APC (0.19.0)

  • APC use new metrics endpoint

  • Allow users to opt out of seeing Docker layer packages

  • Allow adding/removing packages and env variables at the same time

  • Added support for restarting an app via manifest

  • Added the ability to add log drains to an app via manifest

  • Improve APC help text for –basic-ldap

  • Added 'apc app deploy' messages about stopped apps

Web Console

  • Adds Crowd Auth Support

  • Adds LDAP Auth Support

  • Updates Login Page Visuals

  • Adds Useful Info to Services Table

  • Updates visuals on staging page

  • Updates visuals on stager details page

  • Use standard table component on job logs

  • Adds useful info to package details

  • Improves spinner behavior in multiple places

  • Adds/Updates button styles globally

  • Adds Instance Manager details page

  • Uses new endpoint for job metrics (increased performance)

  • Adds metrics to jobs table

  • Releases dashboard page

  • Adds linking from dashboard page to jobs page

  • Improved Error Handling on Cluster Page

  • Client Side Fix for Services name API bug

  • Adds link to windows APC download

  • Adds actual and expected instance count to job table, dashboard

  • Fix visual bug on namespace navigator component

  • Fix visual bug on namespace health indicator

  • Fix dropdown bug

  • Fix filter attribute visual height bug

  • Fixes missing job type bug (and related)

  • Adds depth of element support (shadow depth)

  • Adds release tooling tasks to Gulp

  • Finishes stores refactors (removes metrics store - now unused)

  • Adds Metrics Adapters

  • Removes unused Collection Store

Orchestrator (0.2.2)

  • Changed the default behavior of the deploy command to skip updating the release in order to prevent accidental cluster upgrades

  • Display chef-client output in the case of a chef-client failure

  • Display inventory path in error message for vSphere


  • Added support for installing the Windows 64-bit APC client

  • Added documentation to Working with Staging for debugging stagers using supported command line options

  • Updated Working with Manifests to include support for declaring optional ports and job log drains

  • Added use case explanation to the IP-Manager topic (customer request)

  • Added requirements for using Docker registries

  • Updated Docker connect example

  • Updated trial image installation documentation using VMware Fusion to include the --basic auth requirement

  • Miscellaneous minor typo / broken link fixes

  • Added 428e to the release notes

Jul 08, 2015

Cluster (428e)

  • Cluster changes
    • Evaluate scheduling tags from policy during job compliance check

    • Support for LDAP basic auth

    • Add STAGER_DEBUG environment variable to packages

    • Add full FQN to metrics

    • Allow stagehand plugins to be deleted

    • Remove unused auth types

    • Remove "Act As" functionality from policy

    • Add API hooks for audit entry logging

    • Return a list of enabled identity providers from auth server

    • Statically link OpenSSL

    • Allow identity providers to be explicitly enabled and disabled

APC (0.18.0)

  • Allow users to authenticate using LDAP basic bind credentials.

  • Improvements to help messaging.

Web Console

  • FQN removed from sub-header

  • Fixed issue with text overflowing table column width; values now truncated and elided.

Orchestrator (0.2.2)

  • When a cluster is running a version of Apcera HCOS that is older than the newest promoted release, Orchestrator 0.2.2 does not automatically update the cluster unless specified by CLI option.

  • Display chef-client output in the case of a chef-client failure.

  • Display inventory path in error message for vSphere

  • The location of Kerberos settings in the cluster.conf file was changed.

  • Tuned some postgres parameters related to auto_vacuum (cleanup of disk usage)

  • Updates to list of approved PGP keys for signing APT repositories.

  • Network interfaces without any flags could cause an error in chef parsing of the system configuration.

  • Changes to the Riak startup timing and retries to resolve a problem with first-time Riak configuration on a new cluster.


  • Bug fixed with VMware Fusion login documentation.

  • Added Docker registry support requirements

  • Fixed Docker connect example

Jun 30, 2015

Cluster (427a)

  • Cluster changes
    • Fixed health manager bugs.

    • Use path on disk for staging coordinator.

    • Added logging to workers WaitGroup.

    • Refactored instance manager cgroup container stopCommands.

    • Removed unnecessary iptables traces.

    • Enhanced API_Server route stats.

    • Updated template handling in policy engine.

    • Implemented job quota API.

    • Removed provider FQN for network-provisioned services.

    • Implemented code to detect permission error in package-manager.

    • Replaced HTML 404 page.

    • Implemented interface to enable switching between Cassandra and PG.

    • Display the output if apc docker run fails to stage a package.

    • Fixed deadlock when closing pgstore.

    • Apc package show includes the name of the staging pipeline that was used.

    • Tge createSpEventRule now checks that JobFQN, ServiceFQN, ProviderFQN are in existence in the JM data store.

    • Handle affinity in job update handler.

    • Force umask to 0022 when starting container.

    • Improved JM job delete behavior when failing to connect to Redis.

    • Raise IM resource limits.

    • Ensure APC's WaitForJobStart helper looks for the proper job version.

    • Stop running containers if job state is stopped, respecting job SeqNo when stopping/updating instances.

    • Simplified services API handlers.

    • Fixed service lookup in 'app create', used to panic if service referenced in the manifest did not exist but was created as a part of command flow.

    • Wait for instances to get updated after binding.

    • If SG doesn't have any instances running don't proceed with provider creation.

    • Only respect gateways in RUNNING state as healthy.

    • Updated the EnsureAllInstanceUpdated logic.

    • Added support for '$' in FQN wildcards.

    • Remove service and provider hard-coded FQNs on binding.

    • When restoring don't delete the old file system if we have keep containers on.

    • Verify GRE Tunnel now retries multiple times.

    • Fiex apc export policy bug.

    • API JSON responses include the http status code.

    • Improved build script robustness.

    • Fixed wait loop handling of "running" status.

    • Fixed an empty-message error returned by pgstore.

    • Abort before we subscribe to hearbeats.

    • Allow user to specify staging pipeline even if we can't detect the runtime.

    • Fixed enforcement of StopTimeout on stop commands.

    • Fixed service gateway backend to archive in postgres and not use the non-existant memstore.

    • Changed DNS server to be recursive NS.

    • Fixed race condition between teardown and restore.

    • Improved Restarting jobs, Removed the RunOnce concept.

    • Improved app run cleanup and test.

    • Updated HM logging.

    • Ensure all outbound actions taken by the HM have a request ID on them.

    • Minor logging updates around seqno in IM and chosen instance in API.

    • Fixed proper evaluation if request ID hasnt been set.

    • Upgraded nginx to 1.8, added the nginx-backtrace module.

    • Added API endpoint for job level metrics (/v1/metrics/jobs).

    • Fixed nil checking on code that uses GetJobByID.

    • Removed memstore.

APC (0.17.3)

  • Fixed naming behavior of packages from 'apc package build'.

  • Fixed Semantic Pipeline detection in 'apc app deploy'.

  • Generate same environment variable when allowing egress for a job, whether done during create or update.

  • Updated apc help docker run to include '-sp' option.

  • Improved readability of .apc configuration file.

  • Restart policy may be configured at command line using the '–restart' and the '–restart-attempts' flags on 'app create' and 'docker run'.

  • Output staging logs in 'docker run' failure cases.

  • Improve failed authentication attempt error messages.

  • Many bug fixes around service handling, including polling improvements.

  • Respect batch mode in 'docker run' provider prompt for volumes.

  • Add command-line option to 'app update' / 'job update' to remove egress.

  • Fix some validation in 'app update' / 'job update'.

  • Fix interactive app name entering in 'app from package'.

Web Console

  • On upgrade, you must clear your browser cache and reload static site URLs.

  • Added data-table component.

  • Use data-table for jobs table (visual and code improvements).

  • Cluster metadata now on card component.

  • Added resource quota line to graphs, where appropriate.

  • Added subheader component to details' pages.

  • Added navigable fqn to subheader component.

  • Fixed some font inconsistencies.

  • Dropdown component iteration (mostly code level).

  • Added job and instance stats cards to dashboard page (dark launched).

  • Fixed bug with editing a policy doc twice in one session.

  • Fixed some visual bugs with package detail.

  • Updated graph styles on cluster and job details pages.

  • Updated page layout html/css to improve velocity, visual appeal, screen size and cross browser compatibility.

  • Introduced tablet breakpoint, with toggle-able main navigation.

  • Implemented more cross browser bug fixes (FF & Chrome).

  • Introduced time selection feature for graphs on cluster/job details.

  • Implemented bug fixes on graphs.

  • Updated 'card' component - visual and code quality.

  • Replace most 'stores' with 'api' to enable more client side performance and allow for spinners on individual elements.

Orchestrator (0.2.1)

  • Added support for release bundles which allow for install with no internet connectivity.

  • Console output is now less verbose. Log files are written into the current working directory.

  • Fixed leaked file descriptor.

  • Updated Installation documentation as of June 29, 2015.

  • Detect on all nodes if network.default_route is missing.

  • Added zabbix test to detect if any filesystems are mounted read-only.

  • Created limit for old log-file retention count and size.

  • Moved the pidfile to the new log directory.

  • Moved the router access logs to /var/log/continuum-router-access-logs and setup Splunk ingestion and log rotation.

  • Added a more informative access log for nginx.

  • Changed the alert trigger for orchestrator-agent.

  • Added support for the mountpoint device name being a list.

  • Updated OpenSSL version guard for nginx.

  • Cleaned up statsd to avoid counting zeroes.


  • Added instructions for installing Vagrant trial images.

  • Added examples to capsules documentation.

  • Added examples to Docker documentation.

  • Added examples for working with services.

  • Added examples for working with providers.

  • Updated APC installation and usage documentation.

  • Added cautionary note for using package.lock.

  • Updated --env-var syntax.

  • Updated documentation for working with logs.

  • Added policy information for defaultNamespace.

  • Product re-naming work (Continuum > Apcera HCOS).

  • Made miscellaneous doc bug fixes (typos, links).

  • Added information that distinguishes the “runner” container user used for job instances vs. the "root" user used for app consoles.

  • Added link to "flapping" definition on support site.

  • Added note about concatenating start command messages.

  • Added note about default value for apps and docker jobs.

  • Added new job restart modes for apps and Docker jobs.

May 13, 2015

Cluster (420h)

  • Cluster changes
    • Made general improvements to error messages and timeout handling

    • Fixed panic while evaluating template

    • Fixed service gateway heartbeat subscription logic

APC (0.17.2)

  • Same as last deploy.

Web Console

  • Same as last deploy 418b.


  • Added Release Notes entry for 418b

  • Added Default Support reference topic (/reference/default-support/)

  • Added --mountpath flag and examples to the NFS section in Working with Services (/tutorials/services/#using-nfs-services)

  • Added content for using NFS volumes with Docker images for persistence (/tutorials/docker/#using-docker-with-nfs-for-persistence)

  • Added content for creating/running Docker images from a private Docker repo (/tutorials/docker/#running-a-docker-image-from-a-private-repository)

  • Added explanation around image size to the AMI trial image setup instructions (/setup/public-ami/)

  • Made minor improvements to the Stager API documentation (/api/stager-api)

Apr 29, 2015

Cluster (418b)

  • Cluster changes
    • Implemented enhancement to only produce valid environment variable names for bindings.

    • Implemented feature for tuning base jobs memory consumption.

    • Added namespace quota on packages count.

    • Added namespace quota on jobs count.

    • Fixed dynamic binding ports becoming stale after IM restart.

    • Fixed SP not being properly deleted when leader job is deleted.

APC (0.0.0)

  • Added '-rm' flag to 'docker run' to remove job after SSH connection is closed.

  • Implemented copy memory constraints from Docker image metadata if present.

  • Fixed FQN parsing in 'package update' command when package name is changed.

  • Fixed Docker WORKDIR handling in interactive mode for 'docker run'.

Web Console

  • Same as last deploy 417c.


  • Enabled load balancing of riak through nginx to allow for better PM fault tolerance.

  • Cleaned up older riak configuration.

  • Performed SSH pubkey maintenance, adding new users.

  • Added a forced localhost entry for /etc/hosts for

  • Removed foreign CPU architectures from airgapped APT config, which was causing incorrect architectures to be installed and fail.

  • Cleaned up raft settings for HM.

  • Fixed incorrect NFS export settings along with other NFS recipe cleanup.


  • Updated OpenStack setup instructions to include links to VDI and QCOW2 images.

  • Updated links to AWS AMI images.

Apr 22, 2015

Cluster (417c)

  • Cluster changes
    • Added policydoc::/ policy (with automatic migration) to control who can create, read, and delete, policy documents themselves.

    • Added micro DNS server inside the Instance Manager to avoid using

    • Implemented error handling for failed placement of staging coordinators.

    • Created an exception to allow some jobs to exist without start commands (Docker OS images).

    • Added improvement to perform AUFS mounts one at a time, rather than all at once.

    • Improved network security between containers and cluster hosts.

    • Added support for alternate Docker registries that are compatible with the Docker Hub API.

    • Fixed cluster egress lockdown by allowing egress only to router IPs.

    • Implemented a verification check for system-specific job tags on create and update.

    • Improved filtering of environment variables written to /etc/profile in IM sshd.

    • Added raft election to the Health Manager.

    • Modularized the Continuum token issuance package.

    • Added policy checks around staging pipeline event rule modifications.

    • Added FQN size limit check.

    • Implemented initial modifications to bring restart configuration to HM/JM/API/APC for jobs.

APC (0.0.0)

  • Added support for alternate Docker registries that are compatible with the Docker Hub API (for 'docker run' and 'docker create' commands).

  • Added '–stop-cmd' and '–stop-timeout' parameters to 'app create', 'app update', 'docker run', and other commands to control how jobs stop.

  • Added optional '–instance' parameter to 'app stop', 'job stop' commands to stop a single instance at a time.

  • Enhanced how SSH environment variables are populated via '/etc/profile' when using the 'caspule connect' command.

  • Added a 'package download' command to download raw package contents.

  • Implemented miscellaneous bug fixes related to the console output.

Web Console

  • Made improvements to the line graph component.

  • Added Cluster Dashboard.

  • Added policy to job details page.


  • Added support for air-gapped installations where only the Orchestrator instance has access to the internet. (Requires Orchestrator version 0.2.0 or later.)

  • Restored behavior that honors DNS servers to be used inside containers when specified in the cluster.conf file. (This may be important in installations where hosts have multiple NICs.) To support this change, the container_dns_ip field in the cluster.conf file can now be an array or comma-separated string of multiple DNS server IPs.

  • Added support for installing Splunk using a trial license.

  • Updated the nfs-server role to enable NFSv4 for use by the new persistence feature in Continuum.


  • Migrated FAQs to the Support site to initiate the Knowledgebase.

  • Created standalone section for "Working with Templates" and added examples for job package templating.

  • Updated the Setup section with mention that the OVA file supports both Fusion and vSphere.

  • Updated APC installation to include an example for using multiple email accounts by setting the APC_HOME environment variable.

  • Created a Release Notes index with reference to the Support site for actual editions.

  • Added syntax example for calling custom stagers.

  • Updated FAQ for Docker private registry support.

  • Added FAQ for default staging pipeline.

Apr 08, 2015

Cluster (415a)

  • Cluster changes
    • Established namespace limits on the number of open files and processes

    • Upgraded the Instance Manager (IM) configuration to support multiple DNS servers

    • Moved the Staging Coordinator to the /apcera namespace to be consistent with other system-provided jobs

    • Implemented a shared container environment to address container linking size limit failures

    • Created a system event that is published to the job log when Continuum fails to resolve an NFS mount

    • Created a system event that is published to the job log when the IM stops an instance of a job

    • Implemented a verification check in the monitoring store before attempting to delete a log entry

APC (0.0.0)

  • Updated service bind/unbind API

  • Added '-n' flag to service bind for binding name

  • Bug fixes

Web Console

  • Improved the line graph component

  • Added features to the Cluster Dashboard

  • Added policy to the Job Details page


  • Updated the Monitoring component to use the PostgreSQL password from either the configuration file (preferred) or the node so that both pre-specified and auto-generated passwords now work


  • Removed quota from policy example on job::\ realm

  • Fixed typo in Service Gateway API section

Sep 17, 2014


  • Cluster changes
    • Removed '/bindings' GET endpoint

    • Added generic service gateway

    • Removed 'permit list' from policy grammar

APC (0.13.0)

  • Updated service bind/unbind API

  • Added '-n' flag to service bind for binding name

  • Bug fixes

Sep 15, 2014


  • Cluster changes
    • Disallow instance updates for capsule and docker jobs

    • Cleanup of default policy

    • Added policy for unbinding and unlinking

    • Enhancements to staging pipeline policy

    • Improvements to built-in service gateways

APC (0.12.8)

  • Added docker volume support

  • The app deploy command now copies environment from old package

  • Added job FQN option to rule create

  • Updated job link/unlink to auto-detect port

  • Improvements to error and help messaging

  • Bug fixes

Web Console

  • Bug fixes

Sep 02, 2014


  • Cluster changes
    • System components publish messages to clients via job logs

    • Improved speed of job health lookups

    • Added policy hooks to lock Docker jobs to a single base image

APC (0.12.7)

  • The job logs and app logs commands now contain log messages published by the system

  • Added a policy export command

  • Simplified the default staging pipeline lookup during app create and app deploy

  • Help text improvements

  • Bug fixes

Aug 20, 2014

APC (0.12.5)

  • The apc target command will now ask to use HTTP if the protocol is not specified and HTTPS is unavailable

  • The job list command omits instance counts for jobs that are not running

  • You can now open egress on app create, or with the --allow-egress flag

  • Added display of job version and UUID to list commands

  • Added job-to-job bindings

  • Improvements to error and help messaging

Web Console

  • Usability enhancements and style updates

Aug 18, 2014


  • Cluster changes
    • Add HTTP binding to Redis and Memcache service gateways

Aug 13, 2014

APC (0.12.4)

  • Network egress is now closed by default

  • Apps or capsules may be created with –allow-egress to retain the previous behavior

  • Improvements to resource recommendations based on policy

  • Bug fixes and improvements to error and help messaging

Aug 07, 2014


  • Cluster changes
    • Disk space limits are enforced for applications

APC (0.12.3)

  • Docker improvements

  • Bug fixes and improvements

Web Console

  • Bug fixes and improvements

Aug 05, 2014


  • Cluster changes
    • Allow SSH to app instances by tagging a app with ssh

    • Improvements to Docker staging and usability

    • Improvements to metrics collection and display

APC (0.12.2)

  • Docker apps created with 'docker run' are now accessible over SSH using the docker connect' command

  • Improved output of 'app stats' and 'job stats'

  • Improved port/route display on 'app show' and 'job show'

  • Bug fixes

  • Improved error messaging

Jul 31, 2014


  • Cluster changes
    • All staging errors are now shown properly, including errors from staging coordinator

    • Temporal checks are now supported in policy, for example, to allow retiring a runtime with a known security problem

    • Support for single-binary Docker images

APC (0.12.1)

  • Add '–depends-on' to 'app create' to allow specification of package dependencies at the command line

  • The 'job instances' command now shows instance uptime

  • Environment variable names added by bindings are now displayed with 'app show'

  • The 'app create' command now accepts multiple routes, comma separated

  • Improvements to parsing and display of network rates

  • Various experience improvements

  • Bug fixes

Web Console

  • Policy editor now keeps changes if file doesn't save

Jul 23, 2014


  • Cluster changes
    • Job staging now respects policy, including resource limits for memory and hard disk

    • Package resolution now uses policy, and "blessing" packages has been removed

APC (0.12.0)

  • Resource limits on new applications are now automatically corrected

  • Various improvements and bug fixes

Jul 16, 2014


  • Cluster changes
    • Security hardening via userspace containers

    • System Redis instances persist their data using an AOF when possible

    • Simplified default cluster policy to a single file

    • Audit logs are collected and persisted

APC (0.11.9)

  • Added parallel staging to 'docker run'

  • Removed 'audit log' command in favor of web console view

  • Bug fixes

Web Console

  • Includes a new display for cluster audit logs

Jun 24, 2014


  • Cluster changes
    • Bugfixes and internal improvements

APC (0.11.7)

  • Added route show command to show which jobs use a given route

Jun 20, 2014


  • Cluster changes
    • Network rate limiting for jobs

    • Redis service gateway without persistence

    • Resource minimums for jobs

APC (0.11.6)

  • Max network throughput can be set on apps, capsules, and other jobs, via the job update command

  • Bugfixes

  • Usability improvements

Web Console

  • Streaming job logs in the job detail view

  • UI bugfixes and improvements

  • Fix Continuum Guide help article links

  • Optimize graphs to not trigger on first load unless in viewport

Jun 12, 2014


  • Cluster changes
    • Job logs API improvements with CORS and SSE

APC (0.11.5)

  • Usability improvements and bug fixes

Web Console

  • Bug fixes and performance improvements

Jun 05, 2014


  • Cluster changes
    • Docker images can use a configurable uid/gid

    • CPU maximum allocation is enforced

    • Fixed bug with TCP router IP address configuration

APC (0.11.4)

  • Added app logs command for tailing recent app and job logs

  • Added global --namespace flag for running a command in a different namespace

  • Added custom user and group flags for docker run

  • CPU limits are now enforced, and 0 (the default) will let a job use all available CPU

  • Bug fixes

  • Error text updates

Web Console

  • Traffic splitting (e.g. blue-green)

  • CPU graph updates

May 28, 2014

APC (0.11.3)

  • Add drain add, drain delete, and drain list commands for configuring syslog drains

  • Error text updates

May 20, 2014


  • Cluster changes
    • Updated API to check if routes are in use, during app create and app update

APC (0.11.2)

  • Add route weight command to dynamically control how much traffic a route receives

  • Add --weight parameter to the route add command

  • Error text updates

  • Help text updates

May 12, 2014

Web Console

  • Policy Editor

May 09, 2014


  • Cluster changes
    • Improved metrics publishing rate

    • Added time-based policy control

APC (0.11.1)

  • Added route add and route delete commands

  • Removed --route-add, --route-delete, --scheme, and --url flags from app update. app update now has --port-add and --port-delete

  • Added --filter-by-package option to app list and job list commands

  • Help text updates

Apr 30, 2014


  • Cluster changes
    • You can now add routes and expose ports to running jobs

    • Internal routes changes

    • Updated Google Oauth endpoint

APC (0.11.0)

  • The provider register command now has a --port flag for specifying job ports, instead of --job-route

  • You can now expose routes and ports on running applications and jobs

  • Bugfixes

Web Console

  • Updated routes and ports display

Apr 23, 2014


  • Cluster changes
    • Java stager now sets JVM options for heap settings, relative to job memory allocation

APC (0.10.11)

  • Skip auto-update check when using --batch

  • Added APC blacklisting of SSL certificates

Apr 17, 2014


  • Cluster changes
    • Rails stager now supports Rails 4, and auto-generates a database.yml file

APC (0.10.10)

  • bugfixes

Apr 09, 2014


  • Cluster changes
    • OpenSSL updates (heartbleed)

    • Fixed a Job API response code bug

APC (0.10.9)

  • Added a new job console command

  • The app deploy command better detects start commands

  • Removed deprecated forms of service bind and unbind commands

Web Console

  • Mac APC installer

  • Bugfixes

Apr 04, 2014


  • Cluster changes
    • Improvements to Google oauth integration

    • Policy API improvements

    • Better calculations for job health

APC (0.10.8)

  • Improved output for the job health command

  • Display service binding environment variable keys during app create

  • Help update for namespace and glossary commands, to list allowed characters

Web Console

  • The / key now focuses on the sidebar filterbox (keyboard shortcut)

  • Detail view slide animation updates

  • Bugfixes

Mar 27, 2014

APC (0.10.7)

  • Bugfix for searching for a staging pipeline during app deploy

Mar 26, 2014


  • Cluster changes
    • Added instrumentation to specify package dependency resolution by policy

    • Fixed PostgreSQL Semantic Pipeline processing around blank prepared statements

    • Updated Service Gateways to better report when deleting a binding, service or provider that has services or bindings on it

APC (0.10.6)

  • Added staging pipeline create and staging pipeline clone commands

  • Updated default binding names in service create and service bind commands

  • The service unbind command now accepts a Job name argument

Mar 19, 2014


  • Cluster changes
    • Improved Jetty support in Java stager

    • Postgres, MySQL, Redis, Memcache, and RabbitMQ service gateways now disallow deletion of services with existing bindings, and deletion of providers with existing services

    • Added limited PATCH support to jobs API

    • Deleted jobs now clean up their associated bindings

APC (0.10.5)

  • APC now seamlessly reauthenticates when using Google Device auth

  • Improvements to APC's autoupdater

Web Console

  • Job health in tooltips

  • Sidebar filtering improvements

  • Misc UX fixes

Mar 13, 2014

APC (0.10.4)

  • The job instances command now shows instance creation time

  • The service show command shows jobs bound to a service

  • The import command now preserves namespaces by default

  • Updated help text

Mar 12, 2014


  • Cluster changes
    • Dynamic job-to-job binding improvements

APC (0.10.3)

  • Correctly assign resources entered via prompt in app create, stager create, and other job creation commands

  • Updated help text

Web Console

  • Start and stop jobs from the detail view

  • Change a job's number of instances (also in the detail view)

  • Esc key closes detail view and footer menus

  • Stager status icons

Mar 06, 2014

APC (0.10.2)

  • Improve new version detection logic in APC's autoupdater

  • Remove Service_Gateway- and Semantic_Pipeline- prefixes in the gateway from package and pipeline from package commands

Mar 05, 2014


  • Cluster changes
    • NFS service gateway

APC (0.10.1)

  • Added job health and job instances commands to see job status

  • The app delete command now deletes its associated service bindings

Web Console

  • Hide graphs for stagers

  • Loading state for detail view graphs

  • Visual indicator for "no jobs" (and for each job type)

Feb 27, 2014

APC (0.10.0)

  • The app list, app show, job list, and job show commands now show the number of running vs. configured instances

  • Added provider show and service show commands

  • Improved errors when attempting to create services, bindings, packages, and providers that already exist

  • Added --timeout flag to job update, app update, and app from package

  • All resources (e.g. disk, memory) can be entered with units (e.g. apc app create demo --memory 1GB --disk 2024MB --network 10Mbps)

  • Updated help text

Web Console

  • new design

  • metrics for running jobs

Feb 24, 2014


  • Cluster changes
    • Fixed issue with metrics reporting for running instances

    • Improved core platform stability

    • Updates to PostgreSQL database provisioning, with improved permission handling

APC (0.9.7)

  • The app create command now allows apps to bind to existing services upon creation, if services are specified in the manifest. No changes are needed for existing manifests.

  • Added fields to the pipeline show command

  • Provide a default route in the app from package command

  • Fixed a bug that affected the job stats and app stats commands. Both now show network traffic as bytes sent and received.

  • Updated help text

Feb 20, 2014


  • Cluster changes
    • Introduced App Manifests

    • Introduced TCP Router

APC (0.9.6)

  • The app create command now looks for continuum.conf manifest files upon app creation

  • Added stager show and staging pipeline show commands

  • Added a service from job command

  • The app create command now takes resource values with units, e.g. apc app create foobar --memory 512MB

  • Default job resources are now 200 ms/s CPU, 256MB memory, 1024MB disk, and 5Mbps network

Feb 18, 2014


  • Cluster changes
    • Fixed IP as a service

    • Fixed NATS reconnect logic

APC (0.9.5)

  • Allow user-supplied parameters for service create

Feb 13, 2014


  • Cluster changes
    • Added the ability for stagers to consume services utilizing application manifests

    • Added the Memcache Service Gateway

    • Added the Memcache Semantic Pipeline

    • Added optimizations to package caching for faster provisioning of instances on newly uploaded packages

    • Added ability for Service Gateways to return a templatized URL in a binding, which will be used for ephemeral credential generation with Semantic Pipelines

    • Added the ability for Service Gateways to generate bindings that link to jobs, and get dynamically updated as the job moves within the cluster

    • Added the ability for Service Gateways to call back into the system and provision resources on behalf of the requesting user

    • Fixed issue with Health Manager handling when scaling down

    • Fixed cleaning up stager jobs after staging has completed

    • Fixed API endpoint for package dependency resolution, to preserve ordering

    • Fixed snapshots and path cleanup to be less aggressive

    • Fixed issue with Go stager not having a temp directory set

    • Fixed issue with the built in stager startup occasionally failing with a "file busy" error

    • Updated the static site stager to not include the local port in a redirect

APC (0.9.4)

  • Added apc app from package command to create an application from any package

  • Introduced preliminary manifest support. See apc manifest for details

  • Fixed bug around loss of terminal state in certain cases

  • Copy updates for apc help

  • APC now allows for flags to be input before required arguments

  • Fixed bug around batch mode and streamlined login flow

  • APC should now correctly return nonzero error codes in error cases

  • Improve usage and help for policy import command

Web Console

  • Deployed new documentation site, Continuum Guide

Jan 28, 2014


  • Cluster changes
    • Ensure that sorting of response objects is more consistent

    • Optimized the audit logs to return the most recent entries

    • Fixed issue with MySQL semantic pipeline, for compatibility with Clustrix

    • Fixed issue with handling of downed applications within the router

Jan 27, 2014


  • Cluster changes
    • Metrics Manager API to Graphite API bridge

    • Improve reliability of HM checks

    • Fix issues with the IP allocator rollover

    • Improvements to Google Authentication

APC (0.9.3)

  • Added policy show command

  • Improve help documentation

  • Improve error messages

Jan 21, 2014


  • Cluster changes
    • Added graphite plugin for the Metrics Manager

    • Improvements to partial qualified name handling

    • Stability improvements for core components

Jan 15, 2014

APC (0.9.2)

  • Fix missing service FQN error when binding services

Jan 14, 2014

APC (0.9.1)

  • APC now looks for APC_HOME environment variable to save settings; defaults to home directory

  • Improve error messages returned when parsing FQNs

  • Better validation around changing namespaces

  • Added policy delete command

Jan 10, 2014


  • Cluster changes
    • Improved APIs for Semantic Pipeline rule management to be more consistent with existing APIs

    • Added the Go stager

APC (0.9.0)

  • Introduced FQNs (fully-qualified names)

  • All commands now filter by current namespace

  • Namespaces can be set and viewed using the namespace command

  • Syntax for package update dependency additions has changed

  • updated app create command to ensure that routes contain url-safe characters

  • Updated rule command parameters to be more consistent with other commands

  • Updated rule create with service/provider handling

Jan 07, 2014


  • Cluster changes
    • Added policy enforcement capabilities to capsule connect command

    • Improved handling and parsing of URLs provided when creating a service provider

Jan 02, 2014


  • Cluster changes
    • Fix for MySQL and Postgres Service Gateway error handling when deleting a database

    • Improvements to internal component startup and policy management

APC (0.8.4)

  • Added --optional flag for update commands to specify that a route is optional

  • Fix for APC upgrade prompting to downgrade on point releases

  • Fixed table formatting issue with the provider list and service list commands

  • Added --blessing-add and --blessing-del to package update command

  • Print command usage errors on stderr

  • Clarified flags used when creating gateways from packages

  • Gateway demote will now only demote one Service Gateway

Dec 20, 2013


  • Cluster changes
    • Integrated the new Service Gateway architecture

APC (0.8.3)

  • Added support for new Service Gateways

  • Renamed service register command to provider register

  • Added provider list command

  • Added provider delete command

  • Added service unbind command

Dec 17, 2013


  • Cluster changes
    • Java stager cleanups, including non-web apps supported, default memory pool size lowered

    • PHP security updates — released v5.5.7 and v5.4.23

Dec 13, 2013


  • Cluster changes
    • Change base policies for internal component permissions to be more consise

    • Added support for authentication through Atlassian Crowd

    • Fixed issue with package dependency resolution allowing multiple references to the same package

    • Expanded support for policy enforcement through packages, jobs, and semantic pipeline rules

APC (0.8.2)

  • Added support for Atlassian Crowd

Dec 10, 2013


  • Cluster changes
    • Expanded policy enforcement to cover Packages and Jobs

APC (0.8.1)

  • Added resource update options (memory, CPU, disk, network) to APC commands for app update, job update and capsule update

  • Improved logic and flow around updating to newer minor releases of APC binary

Web Console

  • Fixed cookie session glitch

Dec 05, 2013


  • Cluster changes
    • Updated MySQL Semantic Pipeline to address issue with MySQL JDBC driver authentication

    • Added support for staging Maven applications with the Java stager

    • Changed default session timeout from 60 minutes to 24 hours

    • Fixed package dependency resolution to handle package pairing to better support multiple package handling around multiple operating systems

APC (0.8)

  • Extended timeout when checking for updates

  • Warn when user attempts to capsule connect in a proxied environment to a non-HTTPS cluster

  • Add option to capsule snapshot for user to specify a client timeout

  • Introduced Google device login to APC, where policy allows

  • Improved output and prompts for support ticket command

  • Increased token timeout from 60 minutes to 24 hours

Dec 03, 2013


  • Cluster changes
    • Added support for Google device login, for end-user authentication

APC (0.7.14)

  • Fixed panics occurring in capsule connect

  • Improved termtable output consistency throughout commands

  • Improved flow when updating package names with package update

Nov 27, 2013


  • Cluster changes
    • Added capturing of request method with audit logs

    • Fixed handling of instance tear-down to include deregistration of its routes

    • Updated Health Monitor to ensure orphaned instances are deregistered with the router

Nov 26, 2013


  • Cluster changes
    • Updated cluster's MySQL to default to utf-8 character set

APC (0.7.13)

  • Added consistent --start-cmd flag across all commands

  • Fixed a bug that caused APC to crash on login if there wasn't a target chosen

  • Show non-evaluated environment variables in app show, job show, and capsule show

Nov 22, 2013


  • Cluster changes
    • Updated Nginx version used in router and added a patch to fix SSL read timeout errors

APC (0.7.12)

  • Fixed a bug in I/O timeouts which was killing long-lived connections

Nov 21, 2013


  • Cluster changes
    • Snapshotting improvements

    • Components now export health endpoints, for monitoring

    • Persistent datastore stability fixes

APC (0.7.11)

  • Added option to support ticket command, which includes the apc.log file generated with the --trace flag

  • Fixed commands for service delete and service bind, and improved docs

  • Updated app update to ensure that route additions and deletions include a url or scheme

  • Add --skip-existing flag to the import command to prevent failures when item to be imported already exists

  • Add restart command for jobs, apps, capsules, and gateways

Nov 19, 2013


  • Cluster changes
    • Added support for fully-qualified naming to Services and Service Gateways

    • Added support for staging Java .war files

    • Added Tomcat package and updated Java packages to include setting JAVA_HOME

    • Added varz-style statistics within the cluster for internal health monitoring

    • Updated snapshot functionality to exclude paths like /tmp and /logs

    • Updated router configuration to better handle upgrade requests for websockets

    • Improvements to how stagers handle streaming output to users

APC (0.7.10)

  • Updated capsule create to takes a --package flag, as well as snapshot flags, for capsule creation

  • Updated package update command can now rename packages

  • The capsule snapshot command now correctly preserves filesystem changes between capsules

  • Java stager now autodetects and stages .war files

Nov 14, 2013


  • Cluster changes
    • Log Collector component powers Audit Logs in Web Console and APC

APC (0.7.9)

  • The package replace command now correctly replaces packages with tarballs

  • Bash and static site stagers correctly autodetect runtimes

  • The service bind and service create commands now scrub bad characters from binding names

  • Jobs that fail staging will now clean up their orphaned packages automatically

  • Use the audit logs command to view audit logs

Web Console

  • Audit Logs (beta)

Nov 12, 2013


  • Cluster changes
    • Fixed configuring egress connectivity defaulting to scheme-based port

APC (0.7.8)

  • Added cntmp info command to show contents of an exported .cntmp file

  • Added job package environment variables to app show and job show

  • Waiting for a job to start now takes into account a job's StartupDuration

  • Improve errors returned when no object is found

Web Console

  • Updated design for sign-in window

  • Performance improvements for Apps page

Nov 07, 2013


  • Cluster changes
    • Fixed issue with removing dependencies from the Staging Coordinator

    • Removed old stager timeout (10 minutes) from the Staging Coordinator

    • Fixed handling of file owners and long link names within package creation in staging

    • Updated Instance Manager to better handle IP allocation collisions

    • Stagers — Added snapshot_path option to apc package build, for specifying a directory within the container space to snapshot on package generation

    • Stagers — Added support for the PHP stager to use nginx+php-fpm over Apache and mod_php

    • Packages — Added php-fpm build script

    • Packages — Added build scripts for Ubuntu 13.04

    • Packages — Documented some baseline specifications for an operating system package

APC (0.7.7)

  • Look for APC_BATCH environment variable, to force all commands into batch mode

  • Error messages are more consistent

  • Environment variables can now be set on App creation

  • Improve App and Package update handling

  • Support for multiple App, Package, Stager, and Staging Pipeline exports

  • The apc capsule connect command now establishes an SSH control socket which can be reused, e.g. for scp

  • Updates can now be done to packages in Ready state

Web Console

  • Updated documentation navigation

  • Bugfixes

Nov 01, 2013


  • Cluster changes
    • Stagers can share package dependencies with the package being staged

    • Packages cannot be deleted if in use by a Job

Oct 31, 2013

Web Console

  • New documentation — Environment Variables, Glossary

  • Updated documentation — Stager API sections for "Uploading Data", "Package Dependencies within a Stager", and "Special Environment Variables"

Oct 30, 2013


  • Cluster changes
    • Improved streaming of stager logs on stager failures

    • Updated Staging Coordinator to better detect stager failures and end the staging process

    • Fixed issue with health tracking on jobs that were flapping and unable to be started

APC (0.7.6)

  • Changelog is displayed after auto-update

  • Automatically clean up packages in system when upload or staging fails

  • Added aliases for capsule stats, job stats and gateway stats

  • Added --silent flag to hide all non-error output when running commands

Web Console

  • HTTPS (beta)

Oct 29, 2013

APC (0.7.5)

  • Added the package build command for building and uploading packages to the cluster

  • optional crash reporting (to Apcera Support) if APC panics for some reason

  • added capsule show command (alias to job show)

Web Console

  • Documentation — Package Configuration doc

Oct 28, 2013


  • Cluster changes
    • Improved responses regarding invalid state transitions for Jobs

    • Added memory limit enforcement for Jobs

    • Improved handling around resource limit enforcement

    • Updated to newer runtime versions for Apache, Nginx, Node.js, Perl, PHP, and Ruby

    • Fixed issue which caused dependency re-resolution for all Job updates

    • Fixed issue with freeing port forwarding reservations on instance teardown

    • Improved log streaming to ensure the final messages from a container are read on teardown

    • Fixed issue with streaming App and Stager output when connected over HTTPS

    • Stagers — fixed updates to dependencies to validate the new dependencies and return an error if a package isn’t available

    • Stagers — updates and improvements to the compiler stager

    • Stagers — improved formatting of output to be more consistent

    • Stagers — updated Node.js stager to skip running npm install if no package.json file is given

    • Sample App repo Readme updates

APC (0.7.4)

  • Don’t require authentication when updating APC

  • Show progress bar while APC is updating

  • Improvements to --trace saving logs and writing when APC is aborted

  • Improved handling of apc target

  • Support for --batch mode, which runs non-interactive

  • Added --staging flag to apc stager create, so you can stage a Stager

  • Improved handling of apc capsule connect

  • Job resources now correctly handle memory and disk units

Web Console

  • Splash screen while app is loading

  • UI updates — Package names in Job view, Units in Job view's Resource table, Footer enhancements

  • Documentation — improved general readability, Stager API doc updates

Oct 23, 2013


  • Cluster changes
    • Policy enforcement on Packages

    • Fix for automatically restarting apps that exit unexpectedly

    • Updated the MySQL services connector, for improved connection management

APC (0.7.3)

  • Implemented --trace flag in APC, which writes all requests to an apc.log file

  • Updated definitions in apc glossary

  • Improved UTF-8-handling for tables

  • The apc version command shows APC's current release version

Web Console

  • Documentation updates — Tutorial, Stager API

  • UI updates — improved /home design, cursor fixes for table views, Services view bugfix

  • https bugfixes

Oct 18, 2013


  • Cluster changes
    • SSL support

    • WebSockets over SSL

    • nginx 404 instead of 500

    • Postgres is available as a System Service

APC (0.7.2)

  • SSL support

  • The apc capsule connect command now runs over SSL

  • Added stager delete command

  • Improved http/https proxy support

Web Console

  • Updated apc instructions

  • Minor UI enhancements

Oct 17, 2013

Web Console

  • Minor UI enhancements

Oct 16, 2013


  • Cluster changes
    • Updated instrumentation for authentication

    • Modified credential and token-handling for on-premise proxies

Web Console

  • Minor UI enhancements

  • Modified credential and token-handling for on-premise proxies