Release Notes (Cloud)
Current Version
1.43.0 (1.43.0-9f1c073a) - September 26th, 2024
- New Feature: Last Push and Pull dates for Templates and Tags.
- New Feature: The Controller’s /api/v1/nodes now return the license expiration date.
- Improvement: Distributions can now be cancelled using the
Cancel Distribution
button in the UI or an API call. - Improvement: Upgraded ETCD to 3.5.16.
- Improvement: Fixed
ignoring unknown vm uuid
false positives showing in the logs. - Improvement: Various log improvements in Controller, Regsitry, and Controller Agent.
- Improvement: Upgraded docker package to use newer ubuntu and include health checks.
- Bug Fix: Preventing orphaning of VM on Node when users Terminate a VM while it’s pulling.
- Bug Fix: The automated agent upgrade process was not joining back to the cluster after self-upgrade.
- Various security patches.
- (Required ETCD Version: 3.5.16)
Previous Versions
1.42.0 (1.42.0-92735895) - June 10th, 2024
- New Feature: Ability to start an Intel VM with a specific VNC Password.
- New Feature: Ability to Search through Templates.
- Improvement: New Sub-Menu for Templates.
- Improvement: New dialogue showing table/list of templates being deleted.
- Improvement: Logging improvements, adding Node ID and other IDs where available.
- Bug Fix: Failures for distribution of templates would show a gear icon and never show progress or failure.
- Bug Fix: Registry package post-install message now shows proper port.
- Bug Fix: Some error log lines for rest_sender.go and client.go were showing as blank.
- Bug Fix: Centralized Logging when HTTPS/TLS is enabled was failing.
- Bug Fix: Unable to view Controller Log streams when auth is enabled.
1.41.1 (1.41.1-ac4f02c4) - March 27th, 2024
- New Feature: Added Health Check Endpoints
- Bug Fix: Arm64 macOS package required Rosetta to install.
1.41.0 (1.41.0-6a102e20) - January 24th, 2024
- New Feature: Ability to set External ID when creating a new VM Instance from the UI.
- New Feature: Ability to join a Node in “Drain Mode”.
- Improvement: VM Instances listing in UI now shows Creation Time + sortable by Creation Time.
1.40.0 (1.40.0-95a63938) - December 6th, 2023
- Bug Fix: Inability to delete three or more selected Templates from the Controller UI.
- This release splits the Combined MacOS Package into two packages. Please see our upgrade guide.
- The standalone macOS Registry package has been modernized to match the controller. You can read more about the changes here.
1.39.0 (1.39.0-d35842e6) - November 20th, 2023
- New Feature: We’ve added a task queue + listing element in the UI showing status of bulk actions like adding groups to nodes, changing node configuration, and will be improved to include other actions over time. It will show on the Controller UI in the bottom right.
- New Feature: The Start Instance now supports setting the Port Forwarding for a VM before start.
- New Feature: The Start Instance now supports setting the VM’s Network Mode.
- New Feature: The Start Instance now supports changing the VM’s disk size.
- Improvement: Clarified UAK expiration functionality with a message on the Edit page.
- Improvement: UAK/TAP now works for HA environments.
- Improvement: Create Instance Node listing is now sorted by hostname.
- Bug Fix: Using the save image API/features with a non-existent tag caused a panic and Instance stuck in “Pushing”.
- Bug Fix: Special characters in root password caused controller -> registry communication errors.
- Bug Fix: Port forwarding rule names now display properly on Instance page.
- Bug Fix: Bulk adding multiple nodes to a node group only sets it for one node. Note: The existing
node_ids
param in the API now only supports a single node ID.
1.38.0 (1.38.0-3989ee24) - Oct 24th, 2023
- New Feature: Resource Permissions for Permission Groups. We have a new Authorization guide you can read over to understand how this feature works.
- At the moment Resource Management will disable Node Groups and break compatibility with CI/CD plugins for Jenkins, etc. We are working to add support soon.
- (Required ETCD Version: 3.5.7)
1.37.0 (1.37.0-4e3ffe71) - Aug 29th, 2023
- Improvement: MacOS Packages are now available for ARM64.
- New Feature: Ability to set
hw.serial
andhw.uuid
through Controller API. - New Feature: Ability to delete a Template from Nodes when deleting from the Controller UI Templates page.
- Improvement: Distribution is now supported for Nodes in Drain Mode.
- Improvement: Agent log improvement: “Version change detected! Stopping runner” -> “Anka version change detected! Stopping runner”
- Improvement: Agent log improvement: Template and Tag are now indicated in the Start VM logs.
- Improvement: All agent Start VM logs will now have the Request ID and Instance ID attached.
- Bug Fix: Some Registry API endpoints were using a mix of FAIL and ERROR in the returned JSON. To make consistent with the Controller API, they all now return FAIL.
- (Required ETCD Version: 3.5.7)
1.36.1 (1.36.1-efbe0727) - July 18th, 2023
- Bug Fix: (macOS package) ANKA_LOCAL_ANKA_REGISTRY ignored when TLS enabled.
- Note: If you’re running the registry on a different IP, and especially in your certs only allow specific IPs, you’ll need to set
ANKA_REGISTRY_LISTEN_ADDRESS
to theIP:PORT
vs the default:PORT
.
- Note: If you’re running the registry on a different IP, and especially in your certs only allow specific IPs, you’ll need to set
- Bug Fix: Create Group modal is gray and unclickable.
- (Standalone Registry: 1.36.1-cced806b)
- (Required ETCD Version: 3.5.7)
1.36.0 (1.36.0-37d289c1) - July 18th, 2023
- Bug Fix: Race condition fixes with VM termination to avoid orphans.
- Bug Fix: Running the Anka docker package, the controller would be unable to write centralized logs.
- Bug Fix: (HA only) Improper handling of second VM start for same Instance leaving orphans.
- Bug Fix: Instance can go to started state even after termination request comes in.
- Improvement: Docker packages have been rewritten to:
- Eliminate .env files and instead include them directly in the
docker-compose.yml
- Eliminate ENVs from Dockerfiles used to generate the images/tags in dockerhub. You will want to review our new docker-compose.yml to see which ENVs are now being set there by default and ensure they are also included in your docker-compose.yml (or .env files). For example, the Registry needs
ANKA_LISTEN_ADDR: :8089
andANKA_BASE_PATH: /mnt/vol
set in the config/yml.
- Eliminate .env files and instead include them directly in the
- Improvement: Registry could (depending on configuration, enabled by default) start an additional un-secure server (on port 8085 by default) that is meant to be used by locally running Controller only. This un-secure server is no longer being ran by the Registry.
- Configurations removed:
ANKA_INTERNAL_LISTEN_ADDR
,ANKA_USE_HTTPS_INTERNAL
- Action required: Please ensure that any configuration using
ANKA_LOCAL_ANKA_REGISTRY
will instead point to the official 8089 port. Your TLS certificates may need to be updated to support a new address.
- Configurations removed:
- Improvement: Registry standalone package for macOS will now automatically restart on failure.
- Improvement: MacOS Controller log directory is now 755 and readable by non-root users.
- Improvement: Removed debug logs like
Could not get instance
. - New Feature: A VM is terminated as an orphan/“zombie” if a Node is reporting it 30 times (or more) and the Controller does not know which Instance to correlate it with. This can now be changed by setting
ANKA_UNKNOWN_VM_THRESHOLD
. - (Standalone Registry: 1.36.0-f6c28a0a)
- (Required ETCD Version: 3.5.7)
1.35.0 (1.35.0-525badf3) - June 20th, 2023
- New Feature: The Controller Agent running on Nodes will now timeout when
anka start
commands take longer than 1m30s to complete. This can be controlled by settingankacluster join --cli-start-timeout {new time here}
. - Improvement: Controller logs will now indicate when a Node has picked up the task to start a VM.
- (Standalone Registry: 1.35.0-16331641)
1.34.0 (1.34.0-4cda29d3) - June 8th, 2023
- Improvement: (AUTH ONLY) The Controller will now exclusively use the Root Token for communication and authentication with the Registry. This means that both the Registry and Controller must have AUTH enabled as well as the same Root Token in their configs. Several ENVs like
ANKA_API_KEY_
,ANKA_CLIENT_
, andANKA_OIDC_
(registry only) are no longer available and necessary. They can be removed from your configuration, but be sure to mirror the AUTH configuration for Root Token and other ENVS from the Controller. - Bug Fix: Azure registry
version upload failed
errors. - Bug Fix: User API Keys button was missing from UI.
- Bug Fix: Azure backed registry not returning entire list of templates from API.
- Bug Fix: Azure Registry failing to clean failed pushes.
- (Standalone Registry: 1.33.0-20746877)
1.33.0 (1.33.0-5465bb03) - March 16th, 2023
- Bug Fix: Getting distribution status returned as null body with OK status.
- Bug Fix: Distribution was blocked for Unhealthy nodes.
- Bug Fix: Distribution (/api/v1/registry/vm/distribute) API response randomly showing
progress: 1
(100%) butstatus: false
indefinitely. - Improvement: New Permissions Management Panel
- Improvement: Code/Explicit Flow support for OIDC.
- ENV
ANKA_OIDC_CACHE_TTL
has been deprecated.
- ENV
- Improvement: ETCD upgraded to 3.5.7.
- New Feature: Kubernetes NGINX Ingress Controller Header Passthrough support.
- (Standalone Registry: 1.33.0-a2d41374)
1.32.0 (1.32.0-c375584c) - Feb 6th, 2023
- New Feature: Drain Mode, allowing users to prevent any new VMs from starting on the node, but not terminate already running VMs and the jobs using them.
- New Feature: Certificate Revocation.
- Bug Fix:
/api/v1/node?id=SOME_ID
would return a different object from/api/v1/node
(without ID). - (Standalone Registry: 1.32.0-81cf056c)
1.31.1 (1.31.1-ab1c787e) - Jan 12th, 2023
- Bug Fix: Registry threw a panic when using OIDC.
- (Standalone Registry: 1.31.1-f997f56d)
1.31.0 (1.31.0-93a0aa2c) - Dec 28th, 2022
- Bug Fix: Streaming logs (from Controller’s Logs page) created high CPU usage on registry.
- Bug Fix: With OIDC enabled, lots of unnecessary logs.
- Bug Fix: Show UAK String button replaced Copy Key String which didn’t work on some browsers.
- Improvement: Controller Agent performance improvement for Anka Nodes with lots of templates.
- New Feature: Support calling group claims from userinfo endpoint with
ANKA_OIDC_USER_INFO
ENV. We have also addedANKA_OIDC_CACHE_TTL
which allows control for how long we cache the returned group list from the endpoint as calls to this endpoint can have a performance impact. - New Feature: Ability to set custom scopes for OIDC with
ANKA_OIDC_SCOPES
ENV. - (Standalone Registry: 1.31.0-5c7ab105)
1.30.1 (1.30.1-299dcb37) - Dec 20th, 2022
- Bug Fix: Anka Cluster/Controller Agent did not support 3.2.0 on intel.
- (Standalone Registry: 1.30.1-3cca9a3a)
1.30.0 (1.30.0-66973267) - Dec 1st, 2022
1.28.0 had a significant change you should be aware of to prevent problems. Please read over the ETCD Compaction and Defragmentation documentation to understand what has changed.
- Bug Fix: Anka 3 users: VNC urls from the Controller Instances page now show the password properly. The URL is also not visible at all if there is no forwarded guest 5900 port for the VM template.
- Bug Fix: Incorrect CPU displayed in dashboard and calculation of default value for –vcpu-override on arm machines.
- Bug Fix: Panic on
ankacluster join
when registry URL in the controller is not configured. - Bug Fix: Anka cloud agent calculates ram incorrectly when doing resource check.
- Bug Fix: Anka cloud agent was terminating pulls while Anka 3 Arm conversion was happening, post pull, and it was reaching a timeout.
- Improvement: Simplified
csr_active_config
by removingforceOff
. - Improvement: Moved
processVMUpdate
logs to log level 1 so the Controller logs are less noisy. - New Feature: Ability to set
--local
and--no-local
when starting VMs through API. - (Standalone Registry: 1.30.0-f0b6506a)
1.29.0 (1.29.0-49077f79) - Oct 3rd, 2022
1.28.0 had a significant change you should be aware of to prevent problems. Please read over the ETCD Compaction and Defragmentation documentation to understand what has changed.
- Bug Fix: Mac package anka-controllerd improperly uses
ANKA_LISTEN_ADDRESS
instead ofANKA_LISTEN_ADDR
ENV. - Bug Fix: Agent tries to install PKG even if download from controller fails.
- Bug Fix: Bad string for custom hardware variable (HVAPIC) when using VM start API endpoint.
- New Feature: Management UI for UAKs
- (Standalone Registry: 1.29.0-8daa18e1)
1.28.0 (1.28.0-347c73ea) - Sep 6th, 2022
This release has a significant change you should be aware of to prevent problems. Please read over the ETCD Compaction and Defragmentation documentation to understand what has changed.
- New Feature: Nodes with an expired Anka license now show “Inactive (Invalid License)” in the Nodes UI and do not accept start VM tasks.
- New Feature: Ability to modifying Registry vm list cache expiration/TTL.
- Improvement: Disabled ETCD defragmentation by default (see the note above) and changed compaction time to 30m.
- Bug Fix: Endless CPU usage increase when the license of a single node in the cluster is removed or expires.
- Bug Fix:
Setting license to
log spam when the only node on a controller is disjoined or license expires. - Bug Fix: Various security related concerns.
- (Standalone Registry: 1.28.0-39c3ded1)
1.27.0 (1.27.0-76c64d00) - Aug 8th, 2022
- New Feature: Start VM Instance API now supports
video_controller
,csr_active_config
, andhvapic
. - Bug Fix: Registry files were being corrupted when the server crashed mid-push.
- Bug Fix: $instance_id does not interpolate for name_template in start vm api.
- Bug Fix: Log rotation/cleaning fails when an unknown file seen in the directory.
- (Standalone Registry: 1.27.0-bc73bfc8)
1.26.0 (1.26.0-7f63ad8a) - Jul 21st, 2022
- Bug Fix: Anka Usage calculations were considering duplicate paths.
- Bug Fix: Registry would not start if central log directory was missing (Azure only).
- Improvement: Upgraded etcd to 3.5.4 to solve data corruption issue.
- (Standalone Registry: 1.26.0-757a6a71)
1.25.0 (1.25.0-b2a027a4) - Jun 23rd, 2022
- Bug Fix: Agent panics when anka CLI start VM returns an error.
- Bug Fix: Disjoining does not clear the node task queue and can cause heavy usage over time for the Controller.
- (Standalone Registry: 1.25.0-c6de18fa)
1.24.1 (1.24.1-41a53e49) - May 25th, 2022
- Bug Fix: Instances would sometimes have a blank tag.
- Bug Fix: The agent will now communicate using
--cacert
instead of--root-cert
(which was removed in Anka 3.x). - Improvement: The registry will now use a memory cache to speed up requests for templates/tags. Any API calls to
/registry/v2/vm
or/registry/vm
benefit from this (the controller api also queries these endpoints when making requests for template or tag information). - (Standalone Registry: 1.24.1-d252eb93)
1.24.0 (1.24.0-28ac9d95) - May 17th, 2022
- New Feature: Start VM Instance API
startup_script
monitoring - Bug Fix: Hanging pulling process is not being terminated by agent.
- Bug Fix: If a previous upgrade has been refused (for appropriate reasons, like requested version and current version are identical), further upgrade requests will fail silently until agent is rebooted.
- Bug Fix: Termination tasks duplicates would end up being queued and use more resources than necessary.
- Bug Fix:
remote error: tls: bad certificate
when the registry has a self-signed certificate. - Bug Fix:
api/v1/node
ram
key/value returned the wrong GB value. - Bug Fix: Pushing to the registry would cause the Templates list in the Controller UI to fail to load for the duration of the push.
- Improvement: More detailed logs for situations when the reserve space threshold is triggered.
- Improvement: Various security patches.
- WARNING: SHA-1 certificate support has been removed for TLS/HTTPS and Cert Auth. You will need to regenerate your certs.
- (Standalone Registry: 1.24.0-261d9df2)
1.23.0 (1.23.0-3523c787) - Apr 19th, 2022
- Bug Fix: VM Capacity reverts after rebooting the node.
- Bug Fix: Retrying a registry push too fast will cause a failure.
- Bug Fix: Controller UI was reporting inaccurate sizes for disk usage.
- Improvement: Older CMD logs are now rotated and cleaned up once the aggregate size of logs reaches 700MB (unless otherwise specified).
- Improvement: Anka 3 (apple processor) Nodes are forced to a capacity of 2 (Apple’s limitation) to prevent confusion.
- New Feature: Request a VM with a specific VLAN & View available VLANs attached to a specific Host/Anka Node (REST API ONLY)
- (Standalone Registry: 1.23.0-f21df6a2)
1.22.0 (1.22.0-5dc750f1) - Jan 10th, 2022
- Bug Fix: Available Node and Template information will now show properly for failed instances.
- New Feature: VM Templates, Instances, and Nodes will all show the architecture (intel or arm) they support.
- (Standalone Registry: 1.22.0-5e54d0d6)
1.21.2 (1.21.2-39f47eee) - Dec 28th, 2021
This is a small patch release for 1.21.0.
- Bug Fix: ARM Agent was not being automatically installed correctly on upgrade (404 package download).
- Bug Fix: The etcd container could not start with certain docker-compose versions due to quotes in etcd.env.
- (Standalone Registry: 1.21.2-b3f4cfe7)
1.21.1 (1.21.1-5aebaf69) - Dec 15th, 2021
This is a small patch release for 1.21.0.
- Bug Fix: Within the docker package’s controller/controller.env, the inline comment for
ANKA_ANKA_REGISTRY
would break URLs if the comment was not removed. - (Standalone Registry: 1.21.1-77c8e66e)
1.21.0 (1.21.0-bcc26b24) - Dec 10th, 2021
- Bug Fix: After a macOS upgrade, the
/var/log/veertu
directory was being removed by Apple and not being recreated. We are not creating this if it doesn’t exist so that logs are not missing when they’re needed. - Bug Fix: Fixed the automated agent upgrade processes that was causing the agent to get stuck on the host, post-controller upgrade.
- Bug Fix: Defragmentation of ETCD was happening more often than configured.
- Improvement: Centralized logs are now rotated, by default, each day. This will prevent logs from growing too large should rotation not be triggered.
- Improvement: Redesigned docker package. This includes changes to the tag we generate in dockerhub (it no longer supports certain flags/options in the Entrypoint). Most customers are already using ENVs, so they should be fine.
- Minimum Registry version required for Controller - 1.19.0
- (Standalone Registry: 1.21.0-ced10c21)
Known issues
Please note that there is a temporary workaround required for a bug that started in versions after 1.18.0 of the Controller/Registry agent which runs on your nodes. All versions of the agent, when noticing that the version of itself does not match the version of the controller, will perform a self-upgrade and restart. The restart seems to be problematic on some setups and leaves a zombie anka_agent process and and Offline status in the controller UI. To work around the bug when upgrading your controller/registry, you’ll need to change the existing steps to include:
- Disjoining all of the nodes first
- Do the controller/registry upgrade
- Run
curl -O http://**{controllerUrlHere}**/pkg/AnkaAgent.pkg && sudo installer -pkg AnkaAgent.pkg -tgt /
on each node to download the new version- And finally join each node back
1.21.0 fixes this issue, so it will not be necessary for future releases. Thanks for your understanding and we are sorry for the inconvenience this causes.
1.20.0 (1.20.0-035872f5) - Nov 10th, 2021
- Bug Fix: Moving networks/IPs now updates the Node IP in the Controller UI/database.
- New Feature: Delete VM Templates from Nodes through the Controller API.
- New Feature: Ability to see what VM Templates are on a node from the Controller API.
- Improvement: The
ankacluster
and Anka Agent now support M1/ARM. - Improvement: Upgraded etcd to 3.5.1.
- Improvement: Various security patches & upgrades for golang.
- Minimum Registry version required for Controller - 1.19.0
- (Standalone Registry: 1.20.0-c83f487d)
Known issues we’re working on fixes for: Please note that there is a temporary workaround required for a bug that started in versions after 1.18.0 of the Controller/Registry agent which runs on your nodes. All versions of the agent, when noticing that the version of itself does not match the version of the controller, will perform a self-upgrade and restart. The restart seems to be problematic on some setups and leaves a zombie anka_agent process and and Offline status in the controller UI. To work around the bug when upgrading your controller/registry, you’ll need to change the existing steps to include:
- Disjoining all of the nodes first
- Do the controller/registry upgrade
- Run
curl -O http://**{controllerUrlHere}**/pkg/AnkaAgent.pkg && sudo installer -pkg AnkaAgent.pkg -tgt /
on each node to download the new version- And finally join each node back
We will be fixing the bug soon. Thanks for your understanding and we are sorry for the inconvenience this causes.
1.19.0 (1.19.0-7c1c1424) - Oct 4th, 2021
- Bug Fix: Registry files (tag files and some .ank) were rarely zeroing out due to bad read/write logic
- Improvement: Controller logs now show with node name instances start on
- (Standalone Registry: 1.19.0-76bca3a)
Known issues we’re working on fixes for: Please note that there is a temporary workaround required for a bug that started in versions after 1.18.0 of the Controller/Registry agent which runs on your nodes. All versions of the agent, when noticing that the version of itself does not match the version of the controller, will perform a self-upgrade and restart. The restart seems to be problematic on some setups and leaves a zombie anka_agent process and and Offline status in the controller UI. To work around the bug when upgrading your controller/registry, you’ll need to change the existing steps to include:
- Disjoining all of the nodes first
- Do the controller/registry upgrade
- Run
curl -O http://**{controllerUrlHere}**/pkg/AnkaAgent.pkg && sudo installer -pkg AnkaAgent.pkg -tgt /
on each node to download the new version- And finally join each node back
We will be fixing the bug soon. Thanks for your understanding and we are sorry for the inconvenience this causes.
1.18.0 (1.18.0-b3bb21bf) - Aug 23rd, 2021
- Bug Fix: Reserved tasks do not get released back to queue
- Bug Fix: Prevent controller crashing due to ETCD related panics and when active ETCD endpoint cannot be reached
- New Feature: Ability to use certs and username/password for controller -> etcd connections | Documentation
- Improvement: Upgrading etcd binaries to 3.4.16
- (Standalone Registry: 1.18.0-04fd94e)
1.17.1 (1.17.1-4aead62f) - July 14th, 2021
- Bug Fix: Chrome based browsers don’t work with root token and SSO/OpenID/Keycloak
- (Standalone Registry: 1.17.1-0966fcd)
1.17.0 (1.17.0-fcb89b75) - June 29th, 2021
- Improvement: The Node UUID is now stored in the Anka Agent plist to avoid it changing between crashes or restarts (be sure to disjoin and join after upgrading)
- Improvement:
ankacluster join
commands will now throw a warning if you have not accepted the Anka EULA - Improvement: The exact commands the agent is running will be output to the logs in the event of an error
- Bug Fix: Binary parameters are ignored if preceding by an unknown parameter
- Bug Fix: Groups are now removed from Nodes if ETCD content is reset/deleted
- Bug Fix:
ankacluster join --vm-stuck-delay
is now functional again - (Standalone Registry: 1.17.0-eb513cc)
1.15.0 (1.15.0-c69e2600) - Apr 6th, 2021
- Bug Fix: Client-side load-balancing with three etcd and/or controller containers throws
etcdserver: mvcc: required revision is a future revision
and causes a random loss of VMs - New Feature:
script_result
object is now returned from Save Image Template API - Bug Fix: The start vm queue code sometimes check tasks from the controller queue. It doesn’t take the tasks but it can create load on the db
- (Standalone Registry: 1.15.0-a88b7bc)
1.14.0 (1.14.0-17620328) - Feb 18th, 2021
- New Feature: A button will now show in the Controller UI allowing you to delete an Offline node (instead of having to issue an API call)
- New Feature: You can now set a specific range of MAC addresses that are assigned to VM instances
- Improvement: The registry now supports ANKA_ environment variables, similar to the controller
- (Standalone Registry: 1.14.0-1e39461)
1.12.0 (1.12.0-65cba643) - Dec 1st, 2020
- Bug Fix: Uninstaller throws exit code of 1
- New Feature: Allow configuration of TLS/SSL Protocols and Ciphers for Controller & Registry
- (Standalone Registry: 1.12.0-1f2c3e1)
1.11.2 (1.11.2-ae2c9036) - Nov 10, 2020
- Bug Fix: Memory leak in controller
- (Standalone Registry: 1.11.2-886e687)
1.11.1 (1.11.1-1df83172) - Oct 13, 2020
- Bug Fix: Cache building / Save Image was causing the node to go offline and terminating any other VMs running on it.
- (Standalone Registry: 1.11.1-b8cecfd)
1.11.0 (1.11.0-59d63cca) - August 31, 2020
- New Feature:
anka-controller
executable from macOS native package now supports--tail {lines}
- Improvement: Removed all unnecessary files from docker .tar.gz archive
- Bug Fix: Pulling a template/tag now calculates the size of the tag properly (instead of only the latest tag for the template)
- Bug Fix: VNC url in Controller UI > Instances page is now a link again
- (Standalone Registry: 1.11.0-d43f91c)
1.10.1 - August 20, 2020
1.10.0 - August 5, 2020
- Bug Fix: Central Logging was preventing nodes from joining
- Bug Fix: Certificates with a space in the Organization (O=) were not supported when trying to set permission groups.
- Bug Fix: Controller crashes if client tries to show permission group after deletion
1.9.1 - June 23, 2020
- Bug Fix: When a user sets a custom img_lib_dir location, the anka agent doesn’t calculate disk usage for the new location
1.9.0 - June 10, 2020
- New Feature: Management of MAC Addresses to prevent rare collision
- New Feature: MAC Address can be specified when creating a VM using the API
1.8.0 - June 16, 2020
- New Feature: Controller REST API now supports modifying the vCPU and RAM for VMs in a stopped state
- New Feature: Controller REST API Start VM Instance object allows for custom key/value metadata
- New Feature: Controller Instances page now allows users to add custom columns (from metadata key/value)
- New Feature: Updated license terms
- Bug Fix: Agent process stays alive despite disjoin
- Bug Fix: Instance sorting in controller UI was not kept on update
- Bug Fix: Sorting in instances page in dashboard is reversed
- Bug Fix: Agent’s HTTP client would sometime fail to parse controller’s responses
Change Log combined package (Mac) version 1.7.1 - Apr 14, 2020
- Bug Fix: Anka agent miscalculates free space to keep when pulling images on the node
Change Log combined package (linux) version 1.7.1 - Apr 14, 2020
- Bug Fix: Anka agent miscalculates free space to keep when pulling images on the node
1.7.0 - Mar 23, 2020
- Bug Fix: ankacluster doesn’t pass max vm count to agent
- Bug Fix: Node upgrade request is sometimes sent more than once
- Bug Fix: Distribute not showing progress
- Bug Fix: Controller moves from ‘enterprise’ to ‘basic’ after uninstall and upgrade anka from 2.2.1 to 2.2.2
- New Feature: Display node storage usage information in the controller dashboard
- New Feature: Show size of templates in the controller dashboard
- New Feature: Show a link to the jenkins job that the vm is running in the controller dashboard
- New Feature: Put prompt while deleting template from the controller dashboard
- New Feature: add reserve disk flag in ankacluster command
- New Feature: Add the option to recover from a crash while uploading files to registry
1.6.0 - Mar 08, 2020
- Bug Fix: Jenkins plugin leaves zombie VMs after upgrade to version 2.204
- Bug Fix: Anka registry pull hangs forever when run from agent if VM has a lot of image files
- New Feature: Handle vm start failures more robustly
- New Feature: Added a fail timeout for template distribution
(Mac) - Version 1.5.4 - Jan 15, 2020#
- Bug Fix: Version 1.5.3 was overwriting /usr/local/bin/anka-controllerd file
(Linux) - Version 1.5.4 - Jan 15, 2020#
Other: Version number change to 1.5.4 to make it match with the mac package