v0.10.0
-
(github-actions[bot]) released this 2024-07-17 11:54:57 +0000 UTC*
Description
The SUSE® Rancher Prime Continuous Delivery v0.10 release brings significant changes, including a redesign that splits components into containers and switches to the controller-runtime framework. The rewritten controllers switch to structured logging. To reduce image size and complexity, v0.10 integrates the gitjob repository, merges various controller codes, and introduces new features such as sharding with node selectors, prometheus metrics, configurable leader election, and worker count for the fleet-controller. The release deprecates the "fleet test" command in favor of “target” and “deploy” with a dry-run option and adds experimental OCI content storage. Bug fixes enhance drift detection, cluster status reporting, and various operational aspects, while automation improvements bolster testing, CI, and the release workflow.
Additions
-
Populate bundle status from bundleDeployment status resources by @rubhanazeem in #2535
-
Adding NodeSelection for Controller Shards by @Tommy12789 in #2505
-
Extend strict TLS mode to
fleet-agent
container by @weyfonk in #2556 -
Add metrics to gitops controller by @Tommy12789 in #2530
-
Use bundleDeployment to store resources in gitRepo status by @rubhanazeem in #2523
-
Adds extraEnv to gitjob and fleet-controller deployments by @0xavi0 in #2493
-
Deprecate "fleet test" command, previously used to debug chart by @manno in #2113
-
Support for Azure DevOps Webhook for gitRepos using SSH URL by @0xavi0 in #2440
-
Keep CRDs when deleting a Bundle by @raulcabello in #2028
-
Add correctDrift to targetCustomization by @raulcabello in #2014
-
Adds option to disable polling for Git repos by @0xavi0 in #2226
-
Only create Content resource if Bundle matches any target by @manno in #2239
-
Removes cluster node status reporting from agent by @0xavi0 in #2190
-
Configurable leader election via chart values by @aruiz14 in #1981
-
Replace ImageScan and Bundle Labels With fleet.yaml Labels by @Tommy12789 in #2360
-
Reduce logging for bundles from cleanup controller by @manno in #2393
-
Delete namespace when a gitRepo is deleted by @rubhanazeem in #2373
-
Add readOnlyRootFilesystem to deployment_gitjob.yaml by @thardeck in #2379
-
Creation of a New Webhook endpoint for Github test Webhook by @Tommy12789 in #2395
-
Add a Parameter to Configure Worker count for fleet-controller by @Tommy12789 in #2430
Bugfixes
-
Ignore status fields during drift detection by @aruiz14 in #2522
-
Fixes error display and improves user experience in git poller by @0xavi0 in #2542
-
Honor Helm’s MaxHistory when using drift correction by @aruiz14 in #2519
-
Remove trailing comma from bootstrap template by @Danil-Grigorev in #2452
-
Do not count deleted bundledeployments in cluster status by @manno in #2454
-
Fixes infinite job generation after ForceSyncGeneration changed by @0xavi0 in #2477
-
GitRepo Initial Checkout when DisablePolling is True by @Tommy12789 in #2469
-
Filter namespaces by the kubernetes.io/metadata.name label by @raulcabello in #2009
-
Make AlphabeticalPolicy string check case insensitive by @aditditto in #2216
-
Ignore not-found errors when deleting bundle deployments by @weyfonk in #2329
-
Add missing serviceName in fleet-agent statefulset by @p-se in #2045
-
Fix gitjob log messages for controller and webhook by @manno in #2367
-
Fix helm deployer ignoring kubeconfig CLI arg by @manno in #2112
-
Return commit for revision instead of HEAD by @0xavi0 in #2410
-
Partially replace cleanup with finalizers by @weyfonk in #2431
-
Moves functionality from gitrepo to gitjob controller by @0xavi0 in #2475
Redesign
Performance
What’s Changed
Click to Expand Details
-
Fix example values.yaml for enabling OCI storage by @bigkevmcd in #2494
-
Add authorize and defaults to gitjob controller by @manno in #2491
-
Security context needs to be on container by @manno in #2506
-
Adds DeleteFunc to bundle’s predicate in gitrepo’s controller by @0xavi0 in #2508
-
Add SecurityContext to job container by @raulcabello in #1860
-
Revert "Add SecurityContext to job container" by @raulcabello in #1872
-
Add SecurityContext to job container by @raulcabello in #1875
-
Add security context to token cleanup job by @puffitos in #1862
-
Move cleanup to a different container by @raulcabello in #1885
-
Fix typo in README.md by @guitarmanvt in #1907
-
Replace wrangler for kubebuilder for generating CRDs by @raulcabello in #1904
-
Remove leftovers from controller-gen migration by @aruiz14 in #1920
-
Move agent management to a separate container by @raulcabello in #1910
-
Change variables to constants in bundle type by @manno in #1986
-
Agent Bundle/Release Cleanup Ignores Non-SUSE® Rancher Prime Continuous Delivery Releases by @manno in #1976
-
Move agentmanagement modules into its folder by @manno in #2032
-
Remove bd requeue, rely on drift correction by @manno in #1985
-
Update helm fork, rename ForceAdopt/Adopt to TakeOwnership by @manno in #2146
-
Propagate bundle and bundledeployment status updates to GitRepo by @weyfonk in #2223
-
Switch fetchNamespace to Get instead of list by @manno in #2205
-
Use portable API for bind mounts in integration tests by @weyfonk in #2229
-
Controller image only install ssh clients by @manno in #2335
-
Remove unused update status code from bundle reconciler by @manno in #2349
-
Simplify shards configuration detection by @weyfonk in #2354
-
metrics: optimize collection calls in reconciliation loops by @p-se in #2406
-
Refactoring CreateJob and UpdateStatus methos in Gitjob Reconciler by @Tommy12789 in #2434
-
Gitjob controller updates GitRepo with generation it is reconciling by @manno in #2451
CI / Automation
Click to Expand Details
-
Improve release api description and default by @thardeck in #2548
-
dev: Make setup-single-cluster capable of setting up multiple clusters by @p-se in #2461
-
Include drift end-to-end tests in CI workflows by @weyfonk in #2531
-
Switch from omg.howdoi.website to sslip.io by @p-se in #2466
-
Fix summary and details tags in bug report template by @sbulage in #2476
-
Increase timeout and waiting for testcontainers by @manno in #2502
-
E2e test to checks status fields by @rubhanazeem in #2503
-
Use the same linter version in GoReleaser as for prs by @thardeck in #2511
-
Fix flaky integration test for git job with polling disabled by @weyfonk in #2514
-
Add tmate debugging info to Nightly E2E workflow by @weyfonk in #1847
-
Reduce replica count in rancher integration tests by @manno in #1840
-
Remove cache action for Go build artifacts by @weyfonk in #1867
-
Prevent debug images from creating GitHub releases by @aruiz14 in #1873
-
Use configmap example for
keepResources
E2E tests by @weyfonk in #1870 -
Update exact SUSE® Rancher Prime Continuous Delivery version when releasing against Rancher by @weyfonk in #1880
-
Use private configmap chart for OCI testing by @weyfonk in #1898
-
Fix flaky smoke test for cluster registration by @weyfonk in #1913
-
Add acceptance tests to AKS and GKE CI workflows by @weyfonk in #1914
-
Make sure that all go.mods are tidy for new prs by @thardeck in #1942
-
Provide release.yaml charts version as array by @thardeck in #1971
-
Switch bundle diff test to use random target namespaces by @manno in #1975
-
Switch upgrade test to fleet-test-data repo by @manno in #1936
-
Remove s390x as runners are not available by @raulcabello in #1995
-
Add config for user/passwords in test infra by @manno in #1987
-
Use a different secret for GITHUB_TOKEN by @aruiz14 in #2016
-
Add initial gitrepo controller integration test by @manno in #2034
-
Infra e2e command uses a published fleet module by @manno in #2137
-
Add dev docs for changing the infra command by @manno in #2147
-
Add gitjob end-to-end tests with SSH key by @weyfonk in #2109
-
Add CI workflow releasing SUSE® Rancher Prime Continuous Delivery charts against a test repo by @weyfonk in #2296
-
Add initial Renovate configuration by @renovate-rancher in #2274
-
Upgrade SUSE® Rancher Prime Continuous Delivery in Rancher only triggered manually by @manno in #2241
-
Replace Drone pipeline with Github actions by @weyfonk in #2168
-
Do not mess up go toolchain during release by @thardeck in #2209
-
Duplicate integrationtests scripts for devs by @manno in #2259
-
Fix "Upgrade SUSE® Rancher Prime Continuous Delivery in Rancher To HEAD" CI by @manno in #2260
-
Fix Helm chart version computation in release workflow by @weyfonk in #2227
-
Bump SUSE® Rancher Prime Continuous Delivery version used within installation documentation to 0.9.3 by @rancherbot in #2338
-
Update dev/README Dockerfile for act container image by @p-se in #2326
-
Publish to docker.io/rancher (@manno)
-
Remove go.mod from infra test setup command by @manno in #2378
-
metrics: Add E2E test for cluster deletion by @p-se in #2371
-
Ignore gosec errors in end-to-end tests infra setup by @weyfonk in #2392
-
Add sharding to SUSE® Rancher Prime Continuous Delivery deployments in CI workflows running end-to-end tests by @weyfonk in #2405
-
dev-scripts: Change apiServerURL for setup-fleet-downstream by @p-se in #2052
-
Improve logging and document issue running
act
in a linked git worktree by @p-se in #2372 -
metrics: make sure metrics work well with sharding by @p-se in #2420
Full Changelog: v0.9.5...v0.10.0
Download
Information retrieved from here