File kubernetes1.29.changes of Package kubernetes1.29
------------------------------------------------------------------- Fri May 17 04:52:20 UTC 2024 - Priyanka Saggu <priyanka.saggu@suse.com> - Update to version 1.29.5: * scheduler: preallocation for NodeToStatusMap * add e2e test when a plugin (in-tree or out-of-tree) return non-existent/illegal nodes, the pod should be scheduled normally if suitable nodes exists * add integration test for pod with pvc has node-affinity to non-existent/existent nodes * test to catch illegal nodeNames returned by PreFilterResult * hotfix when a plugin (in-tree or out-of-tree) return non-existent/illegal nodes, the pod scheduling flow will abort immediately. * kubeadm: fix upgrading external CA cluster to 1.29 * Idempotent transformers in scheduler & kcm * cluster/gce: upgrade cos-97-lts -> cos-109-lts * storage e2e: update hostpath and mock images * add log verbosity to endpoint topology hint loop. * Rename `cluster` to `storage_cluster_id` for apiserver_storage_size_bytes metric ------------------------------------------------------------------- Thu Apr 18 11:11:52 UTC 2024 - Priyanka Saggu <priyanka.saggu@suse.com> - Update to version 1.29.4: * Add envFrom to serviceaccount admission plugin * update debian-base to bookworm-v1.0.2 * [go] Bump images, dependencies and versions to go 1.21.9 and distroless iptables / set-cap * [CVE-2023-45288] .*: bump x/net to v0.23.0 * Keep streams from being set up after closeAllStreamReaders is called * Make websocket heartbeat test timing less flaky * Avoid logging binary junk for frame write failure * Use the right feature gate when updating uncertain volumes * fix panic with SIGSEGV in kubeadm certs check-expiration * retry policy creation for CRD type checking E2E test * Undo double run of the TestWatchSemantics test to avoid hitting timeout * apiserver/storage/cacher: decrease the running time of tests in the cacher package. * Serve watch without resourceVersion from cache and introduce a WatchFromStorageWithoutResourceVersion feature gate to allow serving watch from storage. * Ensure that initial events are sorted for WatchList * fix aggregator path filtering to include / * Filter aggregated apiservice gv * Use DBG=1 for conformance verification * build: use -trimpath in non-DBG mode - Update .spec file to bump go version build requirements: * `BuildRequires: go >= 1.21.9` * ref: https://github.com/kubernetes/kubernetes/blob/v1.29.4/build/dependencies.yaml#L119-L121 ------------------------------------------------------------------- Mon Apr 8 10:55:22 UTC 2024 - Priyanka Saggu <priyanka.saggu@suse.com> - update .spec file to set FORCE_HOST_GO env variable, to force k8s build scripts use the local GOTOOLCHAIN version from the build pipeline (i.e. GOTOOLCHAIN='auto') * ref: https://github.com/kubernetes/kubernetes/blob/9791f0d1f39f3f1e0796add7833c1059325d5098/hack/lib/golang.sh#L507 ------------------------------------------------------------------- Mon Apr 08 05:49:04 UTC 2024 - priyanka.saggu@suse.com - Update to version 1.29.3: * remove --v arg from kubetest2-tester-ginkgo test args * ci: bump kind to latest for kms jobs * [go] Bump images, dependencies and versions to go 1.21.8 / distroless-iptables to v0.4.6 * [CVE-2024-24786] Bump github.com/golang/protobuf v1.5.4, google.golang.org/protobuf v1.33.0 * keep existing PDB conditions when updating status * Validate restartable init container state transition * Test that separation of streams work by using progress notifies * Prevent watch cache starvation, by moving its watch to separate RPC and add a SeparateCacheWatchRPC feature flag to disable this behavior * fix UT failure TestPrintIPAddressList * etcd: Update to version 3.5.12 * build etcd image v3.5.12 * Fix race condition between external-resizer and kubelet * codegen: don't nuke files if only verifying * dra: increase timeout in setupFakeDRADriverGRPCServer to prevent tests to flake. * e2e_node: replace `registry.k8s.io/stress:v1` with agnhost * promote agnhost to v2.47 * fixing issue with GetCPUUsageNanoCores on Windows * Fix cleanup of file subpaths * Fix SELinux e2e tests: add label to observed metrics - Update .spec file to bump go version build requirements: * `BuildRequires: go >= 1.21.18` * ref: https://github.com/kubernetes/kubernetes/blob/v1.29.3/build/dependencies.yaml#L119-L121 ------------------------------------------------------------------- Mon Feb 26 09:02:38 UTC 2024 - priyanka.saggu@suse.com - Update to version 1.29.2: * [go] Bump images, dependencies and versions to go 1.21.7 / setcap/debian-base to bookworm-v1.0.1 and distroless-iptables to v0.4.5 * Revert "kubeadm: fix a bug where the uploaded kubelet configuration in kube-system/kubelet-config ConfigMap does not respect user patch" * kubeadm: better checks if the node is CP during "upgrade node" * test: remove unnecessary skip healthz check from test * bugfix: dont skip reconcile for unchanged policy if last sync failed * Add processStartTime in metrics/slis * Fix deprecated version for pod_scheduling_duration_seconds * flag PersistentVolumeLastPhaseTransitionTime field as beta * If a pvc has an empty storageclass name, don't try to assign a default StorageClass to it. * Fix kubeadm init regression * Drop PendingChanges methods from change trackers, move into UpdateResults * Squash some unnecessarily-chained methods in the change trackers ------------------------------------------------------------------- Thu Feb 22 12:43:30 UTC 2024 - Dominique Leuenberger <dimstar@opensuse.org> - Use %patch -P N instead of deprecated %patchN. ------------------------------------------------------------------- Tue Jan 23 10:07:17 UTC 2024 - priyanka.saggu@suse.com - Update to version 1.29.1: * Revert "Graduate Evented PLEG to Beta" * [go] Bump images, dependencies and versions to go 1.21.6 and distroless iptables to 0.4.4 * Make OpenAPIGetter tolerant of nil * use context for lazy evaluation. * Initialize default attach func regardless of the value of flag * .*: bump golang.org/x/tools to v0.16.1 * fix: 119921 * Keep presence cost to 0 to ensure backward compatibility. * fix(scheduler): fix incorrect loop logic in MultiPoint to avoid a plugin being loaded multiple times * fix falky test pkg/registry/core/persistentvolume/storage TestUpdateStatus * Use a separate function for checking if device was reconstructed * Add tests for checking of uncertain device paths * Fix device uncertain errors on reboot * Fix CSI migration for vSphere volumes * Wire in field dropping for CRDs * Revert "scheduler/nodeaffinity: reduce pod scheduling latency" * Revert "scheduler/NodeUnschedulable: reduce pod scheduling latency" * kube-proxy: Fix etp:Local for externalIPs * Add unit test for Job Controller for panic when PodFailurePolicy is used on 1.28 ------------------------------------------------------------------ Tue Jan 23 10:02:04 UTC 2024 - Priyanka Saggu <priyanka.saggu@suse.com> - initial package for Kubernetes v1.29.0
