v0.32.1

What's Changed

• update license to year 2025 by @a7i in #1593
• [release v0.32.0] bump chart and images by @a7i in #1592
• Automated cherry pick of #1591: bump x/net and x/crypto deps for CVE-2024-45337 and by @a7i in #1600

Full Changelog: v0.32.0...v0.32.1

descheduler-helm-chart-0.32.1

Descheduler for Kubernetes is used to rebalance clusters by evicting pods that can potentially be scheduled on better nodes. In the current implementation, descheduler does not schedule replacement of evicted pods but relies on the default scheduler for that.

descheduler-helm-chart-0.32.0

Descheduler for Kubernetes is used to rebalance clusters by evicting pods that can potentially be scheduled on better nodes. In the current implementation, descheduler does not schedule replacement of evicted pods but relies on the default scheduler for that.

v0.32.0

What's Changed

• descheduler v0.31.0: bump kustomize files by @a7i in #1511
• chore: bump golangci-lint to latest by @a7i in #1513
• descheduler v0.31.0: bump helm chart by @a7i in #1512
• fix: github action Release Charts to have write permissions by @a7i in #1514
• [e2e] abstract common methods by @fanhaouu in #1517
• [DuplicatePods] e2e: build a descheduler image and run the descheduler as a pod by @fanhaouu in #1520
• [FailedPods] e2e: build a descheduler image and run the descheduler as a pod by @fanhaouu in #1521
• [TopologySpreadConstraint] e2e: build a descheduler image and run the descheduler as a pod by @fanhaouu in #1523
• [LeaderElection] e2e: build a descheduler image and run the descheduler as a pod by @fanhaouu in #1522
• Make sure dry runs sees all the resources a normal run would do by @john7doe in #1526
• test: construct e2e deployments through buildTestDeployment by @ingvagabund in #1527
• Node utilization refactoring by @ingvagabund in #1532
• add ignorePodsWithoutPDB option by @john7doe in #1529
• Update Dockerfile - GoLang v 1.22.7 FIX - CVE-2024-34156, CVE-2024-34155 and CVE-2024-34158 by @sagar-18 in #1539
• Update nodes sorting function to respect available resources by @ingvagabund in #1541
• DeschedulerServer: new Apply function for applying configuration by @ingvagabund in #1542
• sortNodesByUsage: drop extended resources as they are already counted in by @ingvagabund in #1546
• nodeutilization: NodeUtilization: make pod utilization extraction configurable by @ingvagabund in #1543
• nodeutilization: evictPodsFromSourceNodes: iterate through existing resources by @ingvagabund in #1544
• nodeutilization: usage2KeysAndValues for constructing a key:value list for InfoS printing resource usage by @ingvagabund in #1549
• bump(golangci-lint)=v1.62.0 by @ingvagabund in #1551
• nodeutilization: separate code responsible for requested resource extraction into a dedicated usage client by @ingvagabund in #1545
• Introduce RequestEviction feature for evicting pods in background (KEP-1397) by @ingvagabund in #1466
• Use actual node resource utilization by consuming kubernetes metrics by @ingvagabund in #1555
• Bump to Go 1.23.3 by @seanmalloy in #1565
• Bump golangci-lint to 1.62.2 by @seanmalloy in #1567
• Bump kind version to v0.25.0 by @seanmalloy in #1568
• Bump golangci-lint timeout from 2m to 5m by @seanmalloy in #1571
• Enable golangci-lint Verbose Output by @seanmalloy in #1572
• bump to kubernetes v0.32.0-rc.2 by @a7i in #1569
• docs: Removing deschedulerPolicy.strategies since it does not exist by @felipewnp in #1570
• Fixed the issue when successfulJobsHistoryLimit and failedJobsHistoryLimit variables are 0. by @icloudnote in #1573
• Bump kind to v0.26.0 by @seanmalloy in #1576
• Improve chart default values by @pipo02mix in #1564
• bump to official kubernetes v0.32.0 deps by @a7i in #1575
• feature(eviction): add event when EvictPod failed by @googs1025 in #1536
• [release v0.32] update kubernetes kind version to 1.32 by @a7i in #1587
• [release v0.32] update docs/readme by @a7i in #1588

New Contributors

• @john7doe made their first contribution in #1526
• @felipewnp made their first contribution in #1570
• @icloudnote made their first contribution in #1573
• @pipo02mix made their first contribution in #1564

Full Changelog: v0.31.0...v0.32.0

descheduler-helm-chart-0.30.2

Descheduler for Kubernetes is used to rebalance clusters by evicting pods that can potentially be scheduled on better nodes. In the current implementation, descheduler does not schedule replacement of evicted pods but relies on the default scheduler for that.

Descheduler v0.31.0

What's Changed

• bump to to 1.22.3 to address CVE-2024-24788 by @a7i in #1408
• Fix the replicas value type for the descheduler helm-chart by @hanyouqing in #1378
• Helm chart - allow 'falsey' value in cmdOption by @omerap12 in #1390
• fix helm's default deschedulerPolicy by @duplabe in #1412
• fix TOC location in Readme by @duplabe in #1413
• use cmd context instead of using context.Background() by @googs1025 in #1416
• fix the issue that the pod anti-filtering rules are not taking effect by @fanhaouu in #1395
• Bump otel semconv to 1.24 by @damemi in #1429
• chore: reduce repetition of go versions by @a7i in #1432
• bump go to 1.22.4 for CVE-2024-24790 and CVE-2024-24789 by @a7i in #1433
• return pod qos in advance by @fanhaouu in #1435
• add validation ut by @googs1025 in #1439
• refactor some methods in e2e test by @googs1025 in #1441
• Automated cherry pick of #1427: helm: upgrade to v0.30.1 by @a7i in #1431
• chore: cleanup duplicated code by @googs1025 in #1438
• fix: return the unmatched cases first by @googs1025 in #1446
• Unit-test: add necessary ownerRef to the pod. by @zhifei92 in #1445
• PodEvictor: refactoring and preparation for eviction requests by @ingvagabund in #1447
• refactor: PodMatchNodeSelector method by @googs1025 in #1450
• return node fit error in advance by @fanhaouu in #1436
• test: generate uid when building a pod by @ingvagabund in #1454
• feat: pod evictor options by @a7i in #1455
• fix: indexer cache error when default evictor is re-initialized by @a7i in #1452
• fix: add info for error return by @googs1025 in #1457
• PodEvictor: turn an exceeded limit into an error by @ingvagabund in #1456
• descheduler_test.go refactoring by @ingvagabund in #1459
• The descheduler supports limiting the total number of pods evicted per rescheduling cycle by @zhifei92 in #1451
• descheduler_test.go: initDescheduler: pass a new ctx with cancel inst… by @ingvagabund in #1460
• pod evictor: make it thread safe by @ingvagabund in #1458
• Update README about maxNoOfPodsToEvictTotal by @zhifei92 in #1463
• refactor: replace k8s.io/utils/pointer with k8s.io/utils/ptr by @eminaktas in #1464
• feat: Add namespace override settings in Helm Chart by @ternbusty in #1444
• fix: helm leader-election typo to resourceNamespace by @a7i in #1443
• helm unit tests by @a7i in #1467
• feat: add init and ephemeral container checks to PodLifeTime by @adammw in #1468
• bump(github.com/golangci/golangci-lint)=v1.59.1 by @ingvagabund in #1473
• e2e: TestTooManyRestarts: run descheduler as a whole instead of a single plugin by @ingvagabund in #1472
• e2e: TestRemoveDuplicates: limit the tested namespace by @ingvagabund in #1471
• KEP-1397: descheduler integration with evacuation API as an alternative to eviction API by @ingvagabund in #1354
• skip eviction when pod creation time is below minPodAge threshold setting by @victorgs in #1475
• fix: minor version parsing in version compatibility check by @a7i in #1430
• e2e: TestLeaderElection: delete the lease and increase the retry period by @ingvagabund in #1469
• bump go to 1.22.5 by @shahar-h in #1476
• [TestTooManyRestarts] e2e: build a descheduler image and run the descheduler as a pod by @ingvagabund in #1474
• docs: Provide OCI annotation for where to find image sources by @sklirg in #1479
• Remove descheduler/v1alpha1 type by @ingvagabund in #1482
• tests: de-duplicate framework handle initialization by @ingvagabund in #1483
• [unit test]: test descheduling limits by @ingvagabund in #1484
• [unit test]: simplify test pod evictor reset by @ingvagabund in #1485
• README: drop v1alpha1 descheduler policy mention by @ingvagabund in #1498
• bump k8s.io libs to v0.31.0 by @a7i in #1496
• chore: upgrade python EOL and action versions by @a7i in #1505
• descheduler v0.31: update e2e test versions by @a7i in #1504
• feat(helm): make securityContext conditional in Deployment and CronJob by @bendikp in #1507
• Plugin args: tag arguments with omitempty to reduce the marshalled json size by @ingvagabund in #1480
• descheduler v0.31: update docs and manifests by @a7i in #1506
• [e2e] no test timeouts, produce the same image tag as in production by @fanhaouu in #1508
• chore: replace github.com/ghodss/yaml with sigs.k8s.io/yaml by @Juneezee in #1510

New Contributors

• @hanyouqing made their first contribution in #1378
• @omerap12 made their first contribution in #1390
• @duplabe made their first contribution in #1412
• @googs1025 made their first contribution in #1416
• @zhifei92 made their first contribution in #1445
• @ternbusty made their first contribution in #1444
• @adammw made their first contribution in #1468
• @shahar-h made their first contribution in #1476
• @sklirg made their first contribution in #1479
• @bendikp made their first contribution in #1507
• @Juneezee made their first contribution in #1510

Full Changelog: v0.30.0...v0.31.0

descheduler-helm-chart-0.31.0

Descheduler for Kubernetes is used to rebalance clusters by evicting pods that can potentially be scheduled on better nodes. In the current implementation, descheduler does not schedule replacement of evicted pods but relies on the default scheduler for that.

Descheduler v0.30.1

What's Changed

• bump to to 1.22.3 to address CVE-2024-24788 by @a7i in #1408
• [release 1.30] update kustomize image tags and helm versions by @a7i in #1405
• Automated cherry pick of #1378: Fix the replicas type for the helm-chart
  #1390: allow 'falsey' value in cmdOption
  #1412: fix helm's default deschedulerPolicy
  #1413: fix TOC location in Readme
  #1416: use cmd context instead of using context.Background()
  #1395: fix the issue that the pod anti-filtering rules are not by @a7i in #1425

Full Changelog: v0.30.0...v0.30.1

descheduler-helm-chart-0.30.1

Descheduler for Kubernetes is used to rebalance clusters by evicting pods that can potentially be scheduled on better nodes. In the current implementation, descheduler does not schedule replacement of evicted pods but relies on the default scheduler for that.

descheduler-helm-chart-0.30.0

Descheduler for Kubernetes is used to rebalance clusters by evicting pods that can potentially be scheduled on better nodes. In the current implementation, descheduler does not schedule replacement of evicted pods but relies on the default scheduler for that.