Changelog v1.33.0

[deckhouse-BOaTswain]

Changelog v1.33.0

Know before update

• Components of the kube-dns module will be restarted.
• Control plane components will be restarted.
• Etcd will be restarted. The quota-backend-bytes parameter added to etcd is calculated depending on control-plane memory capacity.
• Ingress controllers version 0.33+ will be restarted.
• IngressNginx controllers 0.25 and 0.26 are removed. Ingress controller version 1.1 will restart.
• Openvpn and admin panel will be restarted.
• Prometheus will be restarted.

Features

• [candi] Upgrade Yandex Cloud terraform provider to 0.74.0 #1649
• [candi] Added support for Ubuntu 22.04 LTS. #1505
• [candi] Bumped containerd to v1.5.11. #1386
• [candi] Added support for Kubernetes 1.23. #1290
• [candi] Improved candi bundle detection to detect CentOS-based distros. #1173
• [cert-manager] Added support of certificate owner ref on certificate level #1601
• [cert-manager] Added Cloudflare's APIToken support for ClusterIssuer. #1528
• [cloud-provider-aws] Added the ability to configure peering connections to the without-nat and standard layouts. #514
• [cloud-provider-azure] Enabled accelerated networking for new machine-controller-manager instances. #1266
• [cloud-provider-yandex] Changed default platform to standard-v3 for new instances created by machine-controller-manager. #1361
• [cni-cilium] 1. Updated Cilium to v1.11.5
  2. Cilium will no longer terminate host network connections abruptly when Host Policies are in effect: Conntrack table reset on agent restart cilium/cilium#19367 #1620
• [cni-cilium] The new module responsible for providing a network between multiple nodes in a cluster using the cilium. #592
  Without a way to migrate from existing CNIs at this moment.
• [cni-flannel] Bumped flannel to 0.15.1. #1173
• [control-plane-manager] Calculate and add the quota-backend-bytes parameter to etcd. #1389
  Etcd will be restarted. The quota-backend-bytes parameter added to etcd is calculated depending on control-plane memory capacity.
• [deckhouse-controller] Added the edit command for the deckhouse-controller to be able to modify cluster configuration files. #1558
• [dhctl] For new Deckhouse installations images for control-plane (image for pause container, for example) will be used from the Deckhouse registry. #1517
• [extended-monitoring] List objects from the kube-apiserver cache, avoid hitting etcd on each list. It should decrease control plane resource consumption. #1535
• [extended-monitoring] The module is available in the Deckhouse Community Edition and enabled by default. #1488
• [helm] Added deprecated APIs alerts for k8s 1.22 and 1.25 #1461
• [istio] Upgraded to 1.12 or 1.13 and new version control method. #1431
• [keepalived] The module is available in the Deckhouse Enterprise Edition #1488
• [log-shipper] Label filter support for log-shipper. Users will be able to filter log messages based on their metadata labels. #1424
• [namespace-configurator] The module is available in the Deckhouse Community Edition and enabled by default. #1488
• [network-gateway] The module is available in the Deckhouse Enterprise Edition #1488
• [okmeter] The module is available in the Deckhouse Community Edition but requires Okmeter license. #1488
• [openvpn] Hooks are rewritten in Go. #1489
• [openvpn] The module is available in the Deckhouse Community Edition (is in experimental state). #1488
• [openvpn] Added support for UDP protocol. #1432
• [prometheus] Added token auth for Prometheus remote write. #1586
• [prometheus] Prefer to schedule the main and the long-term Prometheus on different nodes. #1551
  The Prometheus main and the Prometheus long-term will restart.
• [prometheus] Grafana 8.5.2 #1536
• [prometheus] Create table with enabled Deckhouse web interfaces on the Grafana home page #1415
• [secret-copier] The module is available in the Deckhouse Community Edition and enabled by default. #1488
• [terraform-manager] Upgrade Yandex Cloud terraform provider to 0.74.0 #1649
• [upmeter] Added probe for Grafana Pods. #1658
• [upmeter] Added probe for OpenVPN Pods. #1658
• [upmeter] Added probe for Longterm Prometheus Pods and basic API response. #1658
• [upmeter] Added probe for Kubernetes Dashboard Pods. #1658
• [upmeter] Added probe for Dex Pods and basic API response. #1658
• [upmeter] Added kubelet metrics check to the probe "monitoring-and-autoscaling/key-metrics-presence". #1658
• [user-authn] Use Gitlab refresh token, call refresh method of any connector only once. #995

Fixes

• [candi] Fix build of the AWS cloud controller manager #1716
• [candi] Remove master node coreFraction setting from YandexClusterConfiguration openapi spec #1617
• [candi] Migrate to cgroupfs on containerd installations. #1386
• [ceph-csi] Fixed missing registry secret. #1733
• [cloud-provider-aws] Fix LoadBalancer type none target group creation. #1741
• [cloud-provider-yandex] Rollback changes to set simple-bridge as default CNI for Yandex. #1582
• [helm] Avoid hook failure on errors #1523
• [ingress-nginx] move to ingressClassName spec #1671
  IngressNginx controllers 0.25 and 0.26 are removed. Ingress controller version 1.1 will restart.
• [ingress-nginx] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Ingress controllers version 0.33+ will be restarted.
• [kube-dns] Updated CoreDNS to v1.9.1 #1537
• [kube-dns] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Components of the kube-dns module will be restarted.
• [linstor] LINSTOR updated to 1.18.1, DRBD module to 9.1.7, linstor-csi to 0.19.0, linstor-scheduler to v0.3.0 #1559
• [log-shipper] Fix integration of the File source with the Elasticsearch destination. #1625
• [log-shipper] Provide structural schemas for log-shipper CRDs #1612
• [log-shipper] Add the rate-limit option to the ClusterLogsDestination CRD. #1498
• [log-shipper] Migrate deprecated elasticsearch fields #1453
• [log-shipper] Send reloading signal to all vector processes in a container on config change. #1430
• [monitoring-kubernetes] Fixes alert UnsupportedContainerRuntimeVersion to support the newest versions of containerd - 1.5.* and 1.6.* and docker 20.*. #1506
• [monitoring-kubernetes] Fix kubelet alerts #1471
• [monitoring-kubernetes] 1. Detect proper version of a ebpf program to run on a given kernel.
  2. If a program fails to compile or attach to the kernel tracing facilities, do not crash the ebpf_exporter. #1120
• [node-local-dns] Updated CoreDNS to v1.9.1 #1537
• [openvpn] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Openvpn and admin panel will be restarted.
• [prometheus] Use new metrics names in alert rules. #1627
• [prometheus] Removed the old prometheus_storage_class_change shell hook which has already been replaced by Go hooks. #1396
• [prometheus] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Prometheus will be restarted.
• [upmeter] Fixed potential error loops in remote write exporter #1579
  If a storage responds with 4xx error, the unaccepted metrics will not be re-sent.
• [upmeter] Added missing User-Agent header to remote write exporter, defined as Upmeter/1.0 (Deckhouse <edition> <version>) #1579
• [upmeter] Fix the correctness of neighbor-via-service probe by using ClusterIP service type. #1549
• [upmeter] UI shows only present data #1405
• [upmeter] Use finite timeout in agent insecure HTTP client #1334
• [upmeter] Fixed slow data loading in #1257
• [user-authn] Change dex-authenticator's port name from http to https #1566

Chore

• [cert-manager] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Components of the cert-manager module will be restarted.
• [chrony] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The chrony module will be restarted.
• [cilium-hubble] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Cilium Hubble will be restarted.
• [cloud-provider-aws] Restarting the csi-controller after cloud config changes. #1571
  The csi-controller will restart.
• [cloud-provider-aws] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The cloud-provider-aws module will be restarted.
• [cloud-provider-azure] Restarting the csi-controller after cloud config changes. #1571
  The csi-controller will restart.
• [cloud-provider-azure] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The cloud-provider-azure module will be restarted.
• [cloud-provider-gcp] Restarting csi-controller after cloud config changes. #1571
  All csi-controllers in all cloud-powered clusters will restart.
• [cloud-provider-gcp] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The cloud-provider-gcp module will be restarted.
• [cloud-provider-openstack] Restarting the csi-controller after cloud config changes. #1571
  The csi-controller will restart.
• [cloud-provider-vsphere] Restarting the csi-controller after cloud config changes. #1571
  The csi-controller will restart.
• [cloud-provider-vsphere] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The cloud-provider-vsphere module will be restarted.
• [cloud-provider-yandex] Restarting the csi-controller after cloud config changes. #1571
  The csi-controller will restart.
• [cloud-provider-yandex] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The cloud-provider-yandex module components will be restarted.
• [cni-cilium] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The operator of the cni-cilium module will be restarted.
• [cni-flannel] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The cni-flannel module will be restarted.
• [cni-simple-bridge] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The cni-simple-bridge module will be restarted.
• [control-plane-manager] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Control plane components will be restarted.
• [dashboard] Dashboard upgrade from 2.2.0 to 2.5.1 #1383
• [dashboard] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Kubernetes Dashboard will be restarted.
• [deckhouse] Updated BASE_ALPINE and BASE_ALPINE_3_15 variables to change versions of the Alpine Linux in base images from 3.12.1 to 3.12.12, and from 3.15 to 3.15.4. #1374
• [descheduler] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The descheduler module will be restarted.
• [docs] Suggest gp3 for bastion instance in AWS-based 'Getting started' #1495
• [extended-monitoring] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The extended-monitoring module will be restarted.
• [flant-integration] Removed unused "flantIntegration.team" field from values schema #1514
• [ingress-nginx] Bump GoGo dependency for the protobuf-exporter to prevent improper input. #1519
• [istio] Anti-affinity for istiod Pods in HA installations. proxyConfig.holdApplicationUntilProxyStarts global flag to guarantee that istio sidecar starts before application container. enableHTTP10 flag to allow HTTP/1.0 requests handling. #1665
• [istio] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The istio module will be restarted.
• [kube-proxy] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12 and from 3.15 to 3.15.4. #1374
  The kube-proxy module will be restarted.
• [local-path-provisioner] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The local-path-provisioner module will be restarted.
• [metallb] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The metallb module will be restarted.
• [monitoring-kubernetes] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Components of the monitoring-kubernetes module will be restarted.
• [network-policy-engine] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The kube-router of the network-policy-engine will be restarted.
• [node-local-dns] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The node-local-dns module will be restarted.
• [node-manager] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The bashible-apiserver, cluster-autoscaler and machine-controller-manager components of the node-manager module will be restarted.
• [operator-prometheus] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The operator-prometheus module will be restarted.
• [pod-reloader] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The pod-reloader module will be restarted.
• [prometheus-metrics-adapter] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The prometheus-metrics-adapter module will be restarted.
• [prometheus-pushgateway] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The prometheus-pushgateway module will be restarted.
• [registrypackages] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
• [snapshot-controller] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The snapshot-controller module will be restarted.
• [terraform-manager] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The terraform-manager module will be restarted.
• [upmeter] Turned off smoke-mini storage classes, so it will not order PVs by default and only serve us network checks. #1635
• [upmeter] Revert the deletion of HTTP handlers for e2e tests #1600
• [upmeter] Switched the logging of upmeter metrics on info level while exporting. #1579
• [upmeter] Renamed current state of groups and probes according to SLA in the Terms of Service #1534
• [upmeter] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The upmeter module will be restarted.
• [user-authn] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  Components of the user-authn module will be restarted. All dex-authenticator Pods will be restarted.
• [vertical-pod-autoscaler] The Alpine Linux version in the base image has been bumped from 3.12.1 to 3.12.12. #1374
  The vertical-pod-autoscaler module will be restarted.

See CHANGELOG v1.33 for more details.

[z9r5]

/changelog

[shvgn]

/changelog

[shvgn]

/changelog

[z9r5]

/changelog

[z9r5]

/changelog

[z9r5]

/changelog

[z9r5]

/changelog