Skip to content

v0.69.0

Choose a tag to compare

View all tags
@vrutkovs vrutkovs released this 22 Apr 11:29
· 339 commits to master since this release
v0.69.0
8e41c83
This commit was signed with the committer’s verified signature.
vrutkovs Vadim Rutkovsky
GPG key ID: FC6176CCEB4A1B10
Verified
Learn about vigilant mode.

  • Dependency: vmoperator: Updated default versions for VM apps to v1.140.0 version

  • Dependency: vmoperator: Updated default versions for VM apps to v1.139.0 version
    Update note 1: deprecated env variables for Prometheus CRs conversion VM_ENABLEDPROMETHEUSCONVERTER_PODMONITOR, VM_ENABLEDPROMETHEUSCONVERTER_SERVICESCRAPE, VM_ENABLEDPROMETHEUSCONVERTER_PROMETHEUSRULE, VM_ENABLEDPROMETHEUSCONVERTER_PROBE, VM_ENABLEDPROMETHEUSCONVERTER_SCRAPECONFIG. Use -controller.disableReconcileFor command-line flag with comma-separated list of controller names, that should be disabled.
    Update note 2: removed operator_prometheus_converter_watch_events_total metric since migration of Prometheus object watchers to controllers made this counter obsolete.
    Update note 3: made controller.prometheusCRD.resyncPeriod command line flag noop, which was relevant to Prometheus object watchers.
    Update note 4: -eula flag is not set by default anymore for VMBackup and VMRestore. To avoid VMCluster/VMSingle rollouts set spec.vmstorage.vmBackup.acceptEula: true for VMCluster and spec.vmBackup.acceptEula: true" for VMSingle and replace it with spec.license during VMSingle/VMCluster upgrade.
    Update note 5: new flag was introduced, so VMAnomaly pods will be recreated. VMAnomaly version should be at least 1.25.
    Update note 6: fixed port name collision in VMSingle and VMCluster, when VMBackup is enabled. To avoid VMSingle/VMCluster pods recreation set VM_USE_OLD_BACKUP_RESTORE_PORT_NAMES environment variable to true.
    Update note 7: the default VMDistributed unauthorizedUserAccessSpec was removed to allow users to decide whether access should be anonymous or authorized. Please check documentation with examples for unauthorized and authorized access and update your CRs accordingly.

  • Dependency: vmoperator: Updated default versions for VM apps to v1.139.0 version

  • Dependency: vmoperator: Updated default versions for VL apps to v1.50.0.

  • FEATURE: helm-converter: new CLI tool that helps with migration from Helm charts to their corresponding Operator Custom Resources (CRs).

  • FEATURE: vmsingle: VMSingle reuses vmagent implementation to allow scraping and relabelling. See #1694

  • FEATURE: vmoperator: perform statefulset pods deletion instead of eviction when maxUnavailable set to 100%, which is important for minimum downtime strategy. See #1706.

  • FEATURE: vmuser: support referencing multiple targets of the same kind at targetRefs[*].crd.objects.

  • FEATURE: vmoperator: prettify reconcile diff in logs, now diff objects show only changed JSON fields.

  • FEATURE: VLCluster, VLSingle, VLAgent, VTCluster, VTSingle, VMCluster, VMAgent, VMAnomaly, VMAlert, VMAlertmanager, VMAuth: add spec.componentVersion as an alternative to spec.clusterVersion. This field also available in all objects deploying pods. See this #1949 issue for details.

  • FEATURE: vmanomaly: add support for settings.retention configuration (ttl and check_interval) in configRawYaml and configSecret. See these docs for details.

  • FEATURE: vmuser: support JWT-based auth.

  • FEATURE: vmagent: support HPA in VMAgent CR and in VMAgent, which is a part of VMDistributed. See #1961.

  • FEATURE: vmagent: VMAgent CRs running in statefulSet mode, including VMAgent components in VMDistributed, now support configuring rolling update strategy behavior. See #1987.

  • FEATURE: vmagent: VMAgent CRs running in DaemonSet mode now support configuring rolling update strategy behavior.

  • FEATURE: vmoperator: Dry-run mode. See #1832.

  • FEATURE: vmanomaly: introduce VMAnomalyConfig CRD to enable dynamic configuration and hot-reload support starting from VMAnomaly version 1.25.0.

  • FEATURE: vmalertmanager: introduce arbitrary fs access feature for VMAlertmanager. See #899

  • FEATURE: vmalertmanagerconfig: add update_message field to SlackConfig. This allows alertmanager to edit the original Slack message in-place when alert status changes instead of sending a new one. Requires alertmanager v0.32.0+. See #2064.

  • BUGFIX: vmbackupmanager: remove deprecated -eula flag from vmbackupmanager and vmrestore container args. See #1319.

  • BUGFIX: vmoperator: VMPodScrape for VLAgent and VMAgent now uses the correct port; previously it used the wrong port and could cause scrape failures. See #1887.

  • BUGFIX: vmdistributed: updated VMAuth config consolidating all VMSelects into a single read and all VMClusters into a single write backend

  • BUGFIX: vmdistributed: fix PVC being owned by StatefulSet and top-level object simultaneously. See #1845.

  • BUGFIX: vmoperator: remove unneeded finalizer from core K8s resources. See #835.

  • BUGFIX: vmdistributed: remove finalizers from VMServiceScrape and VMPodScrape objects, and keep finalizers on VMAgent, VMCluster, and VMAuth when DeletionTimestamp is not empty.

  • BUGFIX: vmsingle and vmagent: previously, ingest-only mode could still mount scrape configuration secrets when relabeling or stream aggregation was configured, which caused unexpected secret mounts and RBAC-related failures; now these secrets are not mounted in ingest-only mode, so deployments start with the expected minimal permissions and avoid related runtime errors. See #1828.

  • BUGFIX: vmoperator: recreate STS if immutable fields changed.

  • BUGFIX: vmoperator: wait for STS deletion in case of recreation without throwing an error.

  • BUGFIX: vmdistributed: ignore VMAuth update/delete operations if controller is disabled.

  • BUGFIX: vmalertmanager: fixed ignored tracing config, when no alertmanagerconfig CRs collected. See #1983.

  • BUGFIX: vmagent: apply scrape class relabellings before job ones. See #1997.

  • BUGFIX: vmanomaly and vmagent: render %SHARD_NUM% placeholder when shard count is greater than 0. See #2001.

  • BUGFIX: vlcluster and vtcluster: do not ignore ExtraStorageNodes for select, when default storage is disabled. See #1910.

  • BUGFIX: vmdistributed: use default stub, when no VMAuth backends are available

  • BUGFIX: vmagent: use volume from spec.volumes as persistent queue volume if its name is persistent-queue-data, previously emptyDir was mounted. See #1677.

  • BUGFIX: vmcluster: use volume from spec.vmstorage.volumes and spec.vmselect.volumes as data and cache volumes if its name is vmstorage-db and vmselect-cachedir respectively. See #784.

  • BUGFIX: vmcluster and vmsingle: renamed vmbackup and vmrestore port names, before they had identical to storage/single port names, which is not allowed.

  • BUGFIX: vmsingle: updated scraping implementation to match vmagent functionality.

Assets 5
Loading