File kubernetes1.21.changes of Package kubernetes1.21
------------------------------------------------------------------- Thu Feb 22 12:45:08 UTC 2024 - Dominique Leuenberger <dimstar@opensuse.org> - Use %patch -P N instead of deprecated %patchN. ------------------------------------------------------------------- Mon Mar 27 09:24:52 UTC 2023 - Robert Munteanu <rombert@apache.org> - Stronger conflicts for completion packages ------------------------------------------------------------------- Mon Mar 27 08:38:51 UTC 2023 - Robert Munteanu <rombert@apache.org> - Split individual completions into separate packages ------------------------------------------------------------------- Thu Mar 2 13:58:30 UTC 2023 - Priyanka Saggu <priyanka.saggu@suse.com> - Update to version 1.21.14: * Release commit for Kubernetes v1.21.14 * Reject proxy requests to 0.0.0.0 as well * etcd-client starts retrying transient errors from the etcd cluster * authn: fix cache mutation by AuthenticatedGroupAdder * Fix OpenAPI loading error caused by empty APIService * Correct event registration for multiple scheduler plugins. * Fix: abort nominating a pod that was already scheduled to a node * cpu manager policy set to none, no one remove container id from container map, lead memory leak * kube-up: use registry.k8s.io for containerd-related jobs * Skip updating Endpoints and EndpointSlice if no relevant fields change * ipvs: remove port opener * iptables: remove port opener * Extract containerID from systemd-style cgroupPath in cri_stats_provider And fix test to generate UUID without dash * update go.mod for golang.org/x/crypto to v0.0.0-20211202192323-5770296d904e * azure_file: try to get secret namespace from ClaimRef * azure_file: add namespace tests for InTree to CSI conversion * Update Go to 1.16.15 * cluster/gce: update konnectivity image tags to v0.0.30 * bump sigs.k8s.io/apiserver-network-proxy/konnectivity-client@v0.0.30 * fix dryrun when ca file exists * fix regression introduced by PR 100320 * bump k8s.io/utils for inotify fix * cronjob_controllerv2: do not filter jobs to be reconciled by labels * fix: remove outdated ipv4 route when the corresponding node is deleted * Updating EndpointSlice strategy to retain node name in topology until field is set * Add PDB selector patch integration test * Revert v1beta1 PodDisruptionBudget select patchStrategy * kubeadm: fix the bug that 'kubeadm init --dry-run --upload-certs' command failed with 'secret not found' error * Mark device as uncertain if unmount device succeeds * add namespace in azurefile volumeid * fix: azurefile volumeid conflict in csi migration * Fix DaemonSet status update retry * Enabling kube-proxy metrics on windows kernel mode * fix: delete non existing disk issue * fix containers order after applying * upgrade sigs.k8s.io/structured-merge-diff/v4 to v4.2.1 * fix: do not delete the lb that does not exist * Fix header mutation race in timeout filter * Don't set sysctl net.ipv4.vs.conn_reuse_mode for kernels >=5.9 * kubeadm: remove the restriction that the ca.crt can only contain one certificate * flake fix: remove the error handler for cronjob integration test * vendor: bump cAdvisor to v0.39.3 * Fix the leak of vSphere client sessions * client-go: Clear the ResourceVersionMatch on paged list calls * Update GCE manifest to use konnectivity v0.0.27 * Update to apiserver-network-proxy v0.0.27 * Improving performance of EndpointSlice controller metrics cache * Mark volume mount as uncertain in case of volume expansion fails * mount-utils: Detect potential stale file handle * Add GetPossiblyMountedVolumesForPod to let kubelet know all volumes were unmounted * Add e2e test for NodeUnstage error cases * Skip creating HNS loadbalancer with empty endpoints * dependencies: Update golang.org/x/net to v0.0.0-20211209124913-491a49abca63 * kubeadm: avoid requiring a CA key during kubeconfig expiration checks * kubeadm: print the CA of kubeconfig files in "check expiration" * kubeadm: validate local etcd certficates during expiration checks * publishing-bot/rules: drop rules for release-1.17 branch * Ensure deletion of pods in queues and cache * kube-scheduler: Increase the duration to expire an assumed pod * Skip check for all topology labels when using system default spreading * workqueue: fix leak in queue preventing objects from being GCed * Update mounter interface in volume manager * Ignore 'wait: no child processes' error when calling mount/umount * Reduce calls to docker from dockershim for stats * Fix concurrent map writes error in kube-apiserver * e2e: node: release-1.21: backport findKubeletServiceName * node: e2e: add test for the checkpoint recovery * devicemanager: checkpoint: support pre-1.20 data * fix: remove VMSS and VMSS instances from SLB backend pool only when necessary * fix: leave the probe path empty for TCP probes * fix: skip instance not found when decoupling vmss from lb * Add test for counting inodes correct with hardlinks * Get inodes and disk usage via pure go ------------------------------------------------------------------- Mon Feb 7 16:21:21 UTC 2022 - Dirk Müller <dmueller@suse.com> - avoid bashism in client-common postinstall script (bsc#1195391) ------------------------------------------------------------------- Wed Dec 08 15:22:24 UTC 2021 - rbrown@suse.com - Update to version 1.21.7: * parameter 'disabled-metrics' is invalid * defer close the rotated log open * Fixed nil pointer dereference * Add tests for checking bind mounts * Check subpath file * Add check for subpaths * Manual cherry pick of kube-openapi changes for release-1.21 Bump kube-openapi against kube-openapi/release-1.21 branch * Fixed unit test SELinux support * Add shortcut for SELinux detection * Don't guess SELinux support on error * Use separate pathSpec for local and remote to properly handle cleaning paths * [go1.16] Update to go1.16.10 * Automated cherry pick of #105122: added keys for structured logging (#105138) * Update debian, debian-iptables, setcap images to pick up CVEs fixes * Fixing how EndpointSlice Mirroring handles Service selector transitions * Fix race condition in logging when request times out * Remove nodes with Cluster Autoscaler taint from LB backends. * Retry detaching FibreChannel volume few times * Regenerate vendor/ * Move error reporting to volume plugins * Retry reading /proc/mounts indifinetly in FC and iSCSI volume reconstruction * ConsistentRead tries 10 times * Bump k8s.io/utils * Fix issue in node status updating VolumeAttached list * Support cgroupv2 in node problem detector test * Update CHANGELOG/CHANGELOG-1.21.md for v1.21.6 * Fix log spam for du failure on pod etc-hosts metrics * Run storage hostpath e2e test client pod as privileged * support more than 100 disk mounts on Windows * Update to go1.16.9 * Fix PreferNominatedNode test * Verifying the auth headers are set for upgraded aggregated API requests * apiserver aggregator upgrade unit test * Aggregator uses the regular transport even if the request requires upgrades * check log directory for restartCount * Copy VolumeSnapshotContent annotations in snapshottable.go test * Bump klog to v2.9.0 * Ignore VMs in vmss delete backend pools * aggregate errors when putting vmss * fix: skip not found nodes when reconciling LB backend address pools * fix: consolidate logs for instance not found error * Remove a duplicate StorageClass creation call * Bump golang.org/x/text to v0.3.6 * Revert 102925: Fix Node Resources plugins score when there are pods with no requests * e2e scheduling priorities: do not reference control loop variable * storege e2etest: Delete restored PVC/Pod in snapshottable * tests: Wait for the network connectivity first * v1.21: Fix test flake in old svc registry * 'New' Event namespace validate failed * Update CHANGELOG/CHANGELOG-1.21.md for v1.21.5 * Refine locking in API Priority and Fairness config controller * kube-controller-manager: properly check generic ephemeral volume feature * Fix null JSON round tripping * Propagate conversion errors * integration test * fix 104329: check for headless before trying to release the ClusterIPs * fix detach disk issue on deleting node * fix: ignore the case when comparing azure tags in service annotation * fix: ignore the case when updating tags * backport metrics grabbing through port-forwarding * remove listx from OWNERS_ALIASES * Only use dualstack if the node and config supports it * Local PV e2e: fix leaked local volumes * Add explicit capability for online volume expansion ------------------------------------------------------------------- Wed Oct 6 15:24:17 UTC 2021 - Dominique Leuenberger <dimstar@opensuse.org> - Bump disk requirements in _constraints to 12GB. Data based on the last successful build consumed storage. ------------------------------------------------------------------- Thu Sep 16 11:26:27 UTC 2021 - rbrown@suse.com - Update to version 1.21.5: * Update to go1.16.8 * legacy-cloud-providers: aws: Add support for consuming web identity credentials * e2e iperf2 change threshold to 10MBps = 80 Mbps * Fix NodeAuthenticator tests in dual stack * Fix the key missing issue for structured log * Fix a small regression in Service updates * Service: Fix semantics for Update wrt allocations * Don't prematurely close reflectors in case of slow initialization in watch based manager * Fix storage class setup in regional_pd.go * backport 104410 to release-1.21 * pkg/kubelet/cm: use SkipFreezeOnSet * vendor: bump runc to 1.0.2 * Switch to non-deprecated timestamppb.Now() * Fix buckets initialization * fix: ensure InstanceShutdownByProviderID return false for creating Azure VMs * fix: skip case sensitivity when checking Azure NSG rules * Keep MakeMountArgSensitive and add a new signature that receives flags * Update the unit tests to handle mountFlags * Add missing interface method in mount_unsupported.go * Pass additional flags to subpath mount to avoid flakes in certain conditions * Update CHANGELOG/CHANGELOG-1.21.md for v1.21.4 * Copy golang license to staging copies * delete stale UDP conntrack entries for loadbalancer IPs * Set idle and readheader timeouts ------------------------------------------------------------------- Wed Aug 25 11:29:02 UTC 2021 - rbrown@suse.com - Update to version 1.21.4: * Avoid spurious calls to update/delete validation * Update to go1.16.7 * Fix metrics reporting for the deprecated watch path * Update configure-helper.sh * Update configure-helper.sh * Update configure-helper.sh * Fix disruptive subPath test failures * Fix: ignore not a VMSS error for VMAS nodes in reconcileBackendPools * feat: Provide IPv6 support for internal load balancer * Update to using apiserver-network-proxy v1.22 * Make CSR cleaner tolerate objects with invalid status.certificate * disable aufs module * Update pd csi driver images to use v1 images (CSINode, CSIDriver, etc) * storage e2e: patch in RBAC rules for secrets * storage e2e: downgrade hostpath driver * storage e2e: disable health check containers * storage e2e: automate hostpath YAML updates, hostpath v1.6.2 * update comments and owners file for pkg/util/removeall * kubelet: do not call RemoveAll on volumes directory for orphaned pods * Fix panic in master upgrade tests * Update CHANGELOG/CHANGELOG-1.21.md for v1.21.3 * Updated to use konnectivity client v0.0.21, and implemented placeholder context * Fix the code is leaking the defaulting between unrelated pod instances * Simplify use of the fake dynamic client * fix: return empty VMAS name if using standalone VM * wait for endpoints to be available * Remove extra zone test * Fix frameworkImpl.extenders being not set * fix: Refresh VM cache when node is not found * fix: cleanup outdated routes * strip prefix 'v' in coredns version if exists * Fix closing of decorated watcher channel on timeout * fix: delete non existing disk issue * let objects without metadata pass through the managedFields admission controller ------------------------------------------------------------------- Tue Jul 20 10:51:54 UTC 2021 - Richard Brown <rbrown@suse.com> - Introduce revert-coredns-image-renaming.patch to correct new upstream behaviour ------------------------------------------------------------------- Mon Jul 19 14:34:08 UTC 2021 - Dominique Leuenberger <dimstar@opensuse.org> - Drop kubeadm-opensuse-corednsimage.patch: fixed upstream. ------------------------------------------------------------------- Fri Jul 16 13:08:53 UTC 2021 - rbrown@suse.com - Update to version 1.21.3: * move upgrade test frameworks closer to Describe * Update setcap image to buster-v2.0.3 * Update debian-iptables image to buster-v1.6.5 * Update debian-base image to buster-v1.8.0 * Update to go1.16.6 * Bump SMD to v4.1.2 to pick up #102749 fix * generate scheduler merge patches on the pod status instead of the full pod * Loadbalancer IngressIP policy should be configured as non-DSR to enable routing mesh by default * Fix race in attachdetach tests * Fix Node Resources plugins score when there are pods with no requests * Do not throw error when we can't get canonical path * Update CHANGELOG/CHANGELOG-1.21.md for v1.21.2 * Remove error wrap from logs * client-go: reduce log level of reflector again * Update to go1.16.5 * feat: remove ephemeral-storage etcd requirement * endpointslicemirroring controller mirror address status * sched: fix a bug that a preemptor pod exists as a phantom * Revert "Cleanup portforward streams after their usage" * serviceOwnsFrontendIP shouldn't report error when the public IP doesn't match * Return UnschedulableAndUnresolvable when looking up volume-related resources returns NotFound error * Return UnschedulableAndUnresolvable instead of Error when failing to lookup pvc or storageclass in VolumeZone plugin * use subpath for coredns only for default repository * (scheduler e2e) Create balanced pods in parallel * Fix VolumeAttachment garbage collection for migrated PVs * fix error of setting negative value for containerLogMaxSize * Update setcap to buster-v2.0.1 and add setcap to dependencies.yaml * kubeadm: remove e2e test for ClusterStatus * Update etcd image revision * Update debian-base to buster-v1.7.0 * Update debian-iptables to buster-v1.6.1 * Upgrade konnectivity-client for GRPC connection fixes * Respect annotation size limit for SSA last-applied. * Remove unnecessary quotes from get-kube scripts * Fix expired unit test certs * Set cgroups via opencontainer * vendor: bump runc to rc95 * pkg/kubelet/nodeshutdown/systemd: fix for dbus 5.0.4 * kubelet: reuse manager * Azure: avoid setting cached Sku when updating VMSS and VMSS instances * Ref counting is only applicable to Remote endpoints * Make watch order conformance test reliable * Add missing clientset to EBS storage e2e test * Update CHANGELOG/CHANGELOG-1.21.md for v1.21.1 * fixed wrong warning in kube-proxy regarding topology aware hints * fix removing pods from podTopologyHints mapping * Update cos-gpu-installer image * fix: avoid nil-pointer panic when checking the frontend IP configuration * Add jitter to lease controller * Avoid caching the VMSS instances whose network profile is nil * fix: not tagging static public IP * Use CSI driver to determine unique name for migrated in-tree plugins * chunk target operatation for aws targetGroup * Fix watchForLockfileContention memory leak ------------------------------------------------------------------- Tue May 25 09:03:46 UTC 2021 - Richard Brown <rbrown@suse.com> - Toggle ip4/ipv6 forwarding sysctls to ensure all devices are refreshed [boo#1186125] ------------------------------------------------------------------- Thu May 20 13:19:09 UTC 2021 - Richard Brown <rbrown@suse.com> - Restore 90-kubeadm.conf, still needed [boo#1186125] ------------------------------------------------------------------- Wed May 19 11:28:32 UTC 2021 - Richard Brown <rbrown@suse.com> - Stop setting sysctls on kubelet start with `sysctl -a --system`. Instead use more specific sysctl calls in kubelet service start. [boo#1186125] ------------------------------------------------------------------- Wed May 19 11:17:58 UTC 2021 - rbrown@suse.com - Update to version 1.21.1: * [go1.16] Update to go1.16.4 * Automated cherry pick of #101377: Fix validation in kubectl create ingress (#101426) * Update tests to use agnhost 2.32 * Fix EndpointSlice describe panic when an Endpoint doesn't have zone * Add test create service with ns * Set namespace when using kubectl create service * Update pkg/volume/azure_file/azure_provision.go * Normalize share name to not include capital letters * Extend pod start timeout to 5min for storage subpath configmap test * Fix cleanupMountpoint issue for Windows * fix: set "host is down" as corrupted mount * no watch endpointslice in userpace mode * Ensure service deleted when the Azure resource group has been deleted * Update Makefile * Update dependencies.yaml * update debian-base to buster-v1.5.0 for CVEs * fix TestMutatingWebhookResetsInvalidManagedFields flakes * Updating EndpointSlice validation to match Endpoints validation * fix: azure file namespace issue in csi translation * pkg/kubelet: improve the node informer sync check * Remove Limits from scheduling e2e balanced pod resources * Additional CVE-2021-3121 fix * Fix RBAC of generic ephemeral volumes controller * move filesystem resize code to kubernetes/mount-utils and add need resize function * Fix panic when killing container fails * Fix startupProbe behaviour changed * Fix test * [go1.16] Update to go1.16.3 * Fix Job describe for completion mode * Upgrade kustomize-in-kubectl to v4.1.2 * exec test should not run in Parallel as feature gate is not locked * respect ExecProbeTimeout * apf: exempt probes /healthz /livez /readyz * DelegatingAuthenticationOptions TokenReview request timeout * add node handlers to the metaproxier * Updating Topology Aware Hints to support "Auto" value for annotation * list pod list once to avoid timeout * gce configure containerd default_runtime_name * Update the containerd runtime binary path - gce * Cleanup portforward streams after their usage * azure: fix node public IP not able to fetch issues from IMDS * Update CHANGELOG/CHANGELOG-1.21.md for v1.21.0 * e2e test with hostNetwork pods can't run in parallel * do not tag user created public IPs ------------------------------------------------------------------- Mon May 17 12:50:21 UTC 2021 - Richard Brown <rbrown@suse.com> - Revert remove workaround in 10-kubeadm.conf - still needed [boo#1186125] ------------------------------------------------------------------- Wed May 5 15:49:22 UTC 2021 - Anthony Rabbito <hello@anthonyrabbito.com> - Remove workaround in 10-kubeadm.conf for boo#1171770 causing sysctl values to be reset ------------------------------------------------------------------- Mon Apr 26 09:53:12 UTC 2021 - Richard Brown <rbrown@suse.com> - Introduce kubeadm-opensuse-corednsimage.patch to correct new coredns image name reference introduced in kubernetes 1.21 ------------------------------------------------------------------- Wed Apr 14 14:54:52 UTC 2021 - Richard Brown <rbrown@suse.com> - Rebase opensuse-version-checks.patch ------------------------------------------------------------------- Tue Apr 13 13:24:22 UTC 2021 - Richard Brown <rbrown@suse.com> - Initial Package
