v1.25.X

Upgrade Notice

Before upgrading from earlier releases, be sure to read the Kubernetes Urgent Upgrade Notes.

Release v1.25.12+k3s1

This release updates Kubernetes to v1.25.12, and fixes a number of issues.
​ For more details on what’s new, see the Kubernetes release notes. ​

Changes since v1.25.11+k3s1:

​

• Remove file_windows.go (#7856)
• Fix code spell check (#7860)
• Allow k3s to customize apiServerPort on helm-controller (#7873)
• Check if we are on ipv4, ipv6 or dualStack when doing tailscale (#7883)
• Support setting control server URL for Tailscale. (#7894)
• S3 and Startup tests (#7886)
• Fix rootless node password (#7900)
• Backports for 2023-07 release (#7909)
  • Resolved an issue that caused agents joined with kubeadm-style bootstrap tokens to fail to rejoin the cluster when their node object is deleted.
  • The k3s certificate rotate-ca command now supports the data-dir flag.
• Adding cli to custom klipper helm image (#7915)
  • The default helm-controller job image can now be overridden with the —helm-job-image CLI flag
• Generation of certs and keys for etcd gated if etcd is disabled (#7945)
• Don’t use zgrep in check-config if apparmor profile is enforced (#7954)
• Fix image_scan.sh script and download trivy version (#7950) (#7969)
• Adjust default kubeconfig file permissions (#7984)
• Update to v1.25.12 (#8021) ​

Release v1.25.11+k3s1

This release updates Kubernetes to v1.25.11, and fixes a number of issues.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.10+k3s1:

• Update flannel version (#7649)
• Bump vagrant libvirt with fix for plugin installs (#7659)
• E2E Backports - June (#7705)
  • Shortcircuit commands with version or help flags #7683
  • Add Rotation certification Check, remove func to restart agents #7097
  • E2E: Sudo for RunCmdOnNode #7686
• Add private registry e2e test (#7722)
• VPN integration (#7728)
• Fix spelling test (#7752)
• Remove unused libvirt config (#7758)
• Backport version bumps and bugfixes (#7718)
  • The bundled metrics-server has been bumped to v0.6.3, and now uses only secure TLS ciphers by default.
  • The coredns-custom ConfigMap now allows for *.override sections to be included in the .:53 default server block.
  • The K3s core controllers (supervisor, deploy, and helm) no longer use the admin kubeconfig. This makes it easier to determine from access and audit logs which actions are performed by the system, and which are performed by an administrative user.
  • Bumped klipper-lb image to v0.4.4 to resolve an issue that prevented access to ServiceLB ports from localhost when the Service ExternalTrafficPolicy was set to Local.
  • Make LB image configurable when compiling k3s
  • K3s now allows nodes to join the cluster even if the node password secret cannot be created at the time the node joins. The secret create will be retried in the background. This resolves a potential deadlock created by fail-closed validating webhooks that block secret creation, where the webhook is unavailable until new nodes join the cluster to run the webhook pod.
  • The bundled containerd’s aufs/devmapper/zfs snapshotter plugins have been restored. These were unintentionally omitted when moving containerd back into the k3s multicall binary in the previous release.
  • The embedded helm controller has been bumped to v0.15.0, and now supports creating the chart’s target namespace if it does not exist.
• Add format command on Makefile (#7763)
• Fix logging and cleanup in Tailscale (#7784)
• Update Kubernetes to v1.25.11 (#7788)
• Path normalization affecting kubectl proxy conformance test for /api endpoint (#7818)

Release v1.25.10+k3s1

This release updates Kubernetes to v1.25.10, and fixes a number of issues.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.9+k3s1:

• Ensure that klog verbosity is set to the same level as logrus (#7361)
• Add E2E testing in Drone (#7375)
• Add integration tests for etc-snapshot server flags #7377 (#7378)
• CLI + Config Enhancement (#7404)
  • --Tls-sans now accepts multiple arguments: --tls-sans="foo,bar"
  • Prefer-bundled-bin: true now works properly when set in config.yaml.d files
• Migrate netutil methods into /utils/net.go (#7433)
• Bump Runc + Containerd + Docker for CVE fixes (#7452)
• Bump kube-router version to fix a bug when a port name is used (#7461)
• Kube flags and longhorn storage tests 1.25 (#7466)
• Local-storage: Fix permission (#7473)
• Backport version bumps and bugfixes (#7515)
  • K3s now retries the cluster join operation when receiving a “too many learners” error from etcd. This most frequently occurred when attempting to add multiple servers at the same time.
  • K3s once again supports aarch64 nodes with page size > 4k
  • The packaged Traefik version has been bumped to v2.9.10 / chart 21.2.0
  • K3s now prints a more meaningful error when attempting to run from a filesystem mounted noexec.
  • K3s now exits with a proper error message when the server token uses a bootstrap token id.secret format.
  • Fixed an issue where Addon, HelmChart, and HelmChartConfig CRDs were created without structural schema, allowing the creation of custom resources of these types with invalid content.
  • Servers started with the (experimental) —disable-agent flag no longer attempt to run the tunnel authorizer agent component.
  • Fixed an regression that prevented the pod and cluster egress-selector modes from working properly.
  • K3s now correctly passes through etcd-args to the temporary etcd that is used to extract cluster bootstrap data when restarting managed etcd nodes.
  • K3s now properly handles errors obtaining the current etcd cluster member list when a new server is joining the managed etcd cluster.
  • The embedded kine version has been bumped to v0.10.1. This replaces the legacy lib/pq postgres driver with pgx.
  • The bundled CNI plugins have been upgraded to v1.2.0-k3s1. The bandwidth and firewall plugins are now included in the bundle.
  • The embedded Helm controller now supports authenticating to chart repositories via credentials stored in a Secret, as well as passing repo CAs via ConfigMap.
• Bump containerd/runc to v1.7.1-k3s1/v1.1.7 (#7535)
  • The bundled containerd and runc versions have been bumped to v1.7.1-k3s1/v1.1.7
• Wrap error stating that it is coming from netpol (#7548)
• Add ‘-all’ flag to apply to inactive units (#7574)
• Update to v1.25.10-k3s1 (#7582)

Release v1.25.9+k3s1

This release updates Kubernetes to v1.25.9, and fixes a number of issues.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.8+k3s1:

• Enhance check-config (#7164)
• Remove deprecated nodeSelector label beta.kubernetes.io/os (#6970) (#7121)
• Backport version bumps and bugfixes (#7228)
  • The bundled local-path-provisioner version has been bumped to v0.0.24
  • The bundled runc version has been bumped to v1.1.5
  • The bundled coredns version has been bumped to v1.10.1
  • When using an external datastore, K3s now locks the bootstrap key while creating initial cluster bootstrap data, preventing a race condition when multiple servers attempted to initialize the cluster simultaneously.
  • The client load-balancer that maintains connections to active server nodes now closes connections to servers when they are removed from the cluster. This ensures that agent components immediately reconnect to a current cluster member.
  • Fixed a race condition during cluster reset that could cause the operation to hang and time out.
• Updated kube-router to move the default ACCEPT rule at the end of the chain (#7221)
  • The embedded kube-router controller has been updated to fix a regression that caused traffic from pods to be blocked by any default drop/deny rules present on the host. Users should still confirm that any externally-managed firewall rules explicitly allow traffic to/from pod and service networks, but this returns the old behavior that was relied upon by some users.
• Update klipper lb and helm-controller (#7240)
• Update Kube-router ACCEPT rule insertion and install script to clean rules before start (#7276)
  • The embedded kube-router controller has been updated to fix a regression that caused traffic from pods to be blocked by any default drop/deny rules present on the host. Users should still confirm that any externally-managed firewall rules explicitly allow traffic to/from pod and service networks, but this returns the old behavior that was relied upon by some users.
• Update to v1.25.9-k3s1 (#7283)

Release v1.25.8+k3s1

This release updates Kubernetes to v1.25.8, and fixes a number of issues.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.7+k3s1:

• Update flannel and kube-router (#7061)
• Bump various dependencies for CVEs (#7043)
• Enable dependabot (#7045)
• Wait for kubelet port to be ready before setting (#7064)
  • The agent tunnel authorizer now waits for the kubelet to be ready before reading the kubelet port from the node object.
• Adds a warning about editing to the containerd config.toml file (#7075)
• Improve support for rotating the default self-signed certs (#7079)
  • The k3s certificate rotate-ca checks now support rotating self-signed certificates without the --force option.
• Update to v1.25.8-k3s1 (#7106)
• Update flannel to fix NAT issue with old iptables version (#7138)

Release v1.25.7+k3s1

This release updates Kubernetes to v1.25.7, and fixes a number of issues.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.6+k3s1:

• Add jitter to scheduled snapshots and retry harder on conflicts (#6782)
  • Scheduled etcd snapshots are now offset by a short random delay of up to several seconds. This should prevent multi-server clusters from executing pathological behavior when attempting to simultaneously update the snapshot list ConfigMap. The snapshot controller will also be more persistent in attempting to update the snapshot list.
• Bump cri-dockerd (#6798)
  • The embedded cri-dockerd has been updated to v0.3.1
• Bugfix: do not break cert-manager when pprof is enabled (#6837)
• Wait for cri-dockerd socket (#6853)
• Bump vagrant boxes to fedora37 (#6858)
• Fix cronjob example (#6864)
• Ensure flag type consistency (#6867)
• Consolidate E2E tests (#6887)
• Ignore value conflicts when reencrypting secrets (#6919)
• Use default address family when adding kubernetes service address to SAN list (#6904)
  • The apiserver advertised address and IP SAN entry are now set correctly on clusters that use IPv6 as the default IP family.
• Allow ServiceLB to honor ExternalTrafficPolicy=Local (#6907)
  • ServiceLB now honors the Service’s ExternalTrafficPolicy. When set to Local, the LoadBalancer will only advertise addresses of Nodes with a Pod for the Service, and will not forward traffic to other cluster members.
• Fix issue with servicelb startup failure when validating webhooks block creation (#6916)
  • The embedded cloud controller manager will no longer attempt to unconditionally re-create its namespace and serviceaccount on startup. This resolves an issue that could cause a deadlocked cluster when fail-closed webhooks are in use.
• Backport user-provided CA cert and kubeadm bootstrap token support (#6929)
  • K3s now functions properly when the cluster CA certificates are signed by an existing root or intermediate CA. You can find a sample script for generating such certificates before K3s starts in the github repo at contrib/util/certs.sh.
  • K3s now supports kubeadm style join tokens. k3s token create now creates join token secrets, optionally with a limited TTL.
  • K3s agents joined with an expired or deleted token stay in the cluster using existing client certificates via the NodeAuthorization admission plugin, unless their Node object is deleted from the cluster.
• Fix access to hostNetwork port on NodeIP when egress-selector-mode=agent (#6936)
  • Fixed an issue that would cause the apiserver egress proxy to attempt to use the agent tunnel to connect to service endpoints even in agent or disabled mode.
• Updated flannel version to v0.21.1 (#6915)
• Allow for multiple sets of leader-elected controllers (#6941)
  • Fixed an issue where leader-elected controllers for managed etcd did not run on etcd-only nodes
• Fix etcd and ca-cert rotate issues (#6954)
• Fix ServiceLB dual-stack ingress IP listing (#6987)
  • Resolved an issue with ServiceLB that would cause it to advertise node IPv6 addresses, even if the cluster or service was not enabled for dual-stack operation.
• Bump kine to v0.9.9 (#6975)
  • The embedded kine version has been bumped to v0.9.9. Compaction log messages are now omitted at info level for increased visibility.
• Update to v1.25.7-k3s1 (#7010)

Release v1.25.6+k3s1

This release updates Kubernetes to v1.25.6, and fixes a number of issues.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.5+k3s2:

• Pass through default tls-cipher-suites (#6730)
  • The K3s default cipher suites are now explicitly passed in to kube-apiserver, ensuring that all listeners use these values.
• Bump containerd to v1.6.15-k3s1 (#6735)
  • The embedded containerd version has been bumped to v1.6.15-k3s1
• Bump action/download-artifact to v3 (#6747)
• Backport dependabot/updatecli updates (#6761)
• Fix Drone plugins/docker tag for 32 bit arm (#6768)
• Update to v1.25.6+k3s1 (#6775)

Release v1.25.5+k3s2

This release updates containerd to v1.6.14 to resolve an issue where pods would lose their CNI information when containerd was restarted.

Changes since v1.25.5+k3s1:

• Bump containerd to v1.6.14-k3s1 (#6694)
  • The embedded containerd version has been bumped to v1.6.14-k3s1. This includes a backported fix for containerd/7843 which caused pods to lose their CNI info when containerd was restarted, which in turn caused the kubelet to recreate the pod.

Release v1.25.5+k3s1

⚠️ WARNING

This release is affected by https://github.com/containerd/containerd/issues/7843, which causes the kubelet to restart all pods whenever K3s is restarted. For this reason, we have removed this K3s release from the channel server. Please use v1.25.5+k3s2 instead.

This release updates Kubernetes to v1.25.5, and fixes a number of issues.

Breaking Change: K3s no longer includes swanctl and charon binaries. If you are using the ipsec flannel backend, please ensure that the strongswan swanctl and charon packages are installed on your node before upgrading K3s to this release.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.4+k3s1:

• Fix log for flannelExternalIP use case (#6531)
• Fix Carolines github id (#6464)
• Github CI Updates (#6522)
• Add new prefer-bundled-bin experimental flag (#6420)
  • Added new prefer-bundled-bin flag which force K3s to use its bundle binaries over that of the host tools
• Bump containerd to v1.6.10 (#6512)
  • The embedded containerd version has been updated to v1.6.10-k3s1
• Stage the Traefik charts through k3s-charts (#6519)
• Make rootless settings configurable (#6498)
  • The rootless port-driver, cidr, mtu, enable-ipv6, and disable-host-loopback settings can now be configured via environment variables.
• Remove stuff which belongs in the windows executor implementation (#6517)
• Mark v1.25.4+k3s1 as stable (#6534)
• Add prefer-bundled-bin as an agent flag (#6545)
• Bump klipper-helm and klipper-lb versions (#6549)
  • The embedded Load-Balancer controller image has been bumped to klipper-lb:v0.4.0, which includes support for the LoadBalancerSourceRanges field.
  • The embedded Helm controller image has been bumped to klipper-helm:v0.7.4-build20221121
• Switch from Google Buckets to AWS S3 Buckets (#6497)
• Fix passing AWS creds through Dapper (#6567)
• Fix artifact upload with aws s3 cp (#6568)
• Disable CCM metrics port when legacy CCM functionality is disabled (#6572)
  • The embedded cloud-controller-manager’s metrics listener on port 10258 is now disabled when the --disable-cloud-controller flag is set.
• Sync packaged component Deployment config (#6552)
  • Deployments for K3s packaged components now have consistent upgrade strategy and revisionHistoryLimit settings, and will not override scaling decisions by hardcoding the replica count.
  • The packaged metrics-server has been bumped to v0.6.2
• Mark secrets-encryption flag as GA (#6582)
• Bump k3s root to v0.12.0 and remove strongswan binaries (#6400)
  • The embedded k3s-root version has been bumped to v0.12.0, based on buildroot 2022.08.1.
  • The embedded swanctl and charon binaries have been removed. If you are using the ipsec flannel backend, please ensure that the strongswan swanctl and charon packages are installed on your node before upgrading k3s.
• Update flannel to v0.20.2 (#6588)
• Add ADR for security bumps automation (#6559)
• Update node12->node16 based GH actions (#6593)
• Updating rel docs (#6237)
• Update install.sh to recommend current version of k3s-selinux (#6453)
• Update to v1.25.5-k3s1 (#6622)
• Bump containerd to v1.6.12-k3s1 (#6631)
  • The embedded containerd version has been bumped to v1.6.12
• Preload iptable_filter/ip6table_filter (#6646)

Release v1.25.4+k3s1

This release updates Kubernetes to v1.25.4, and fixes a number of issues.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.3+k3s1:

• Add the gateway parameter in netplan (#6292)
• Bumped dynamiclistener library to v0.3.5 (#6300)
• Update kube-router to v1.5.1 with extra logging (#6345)
• Update maintainers (#6298)
• Bump testing to opensuse Leap 15.4 (#6337)
• Update E2E docs with more info on ubuntu 22.04 (#6316)
• Netpol test for podSelector & ingress (#6247)
• Bump all alpine images to 3.16 (#6334)
• Bump kine to v0.9.6 / sqlite3 v3.39.2 (CVE-2022-35737) (#6317)
• Add hardened cluster and upgrade tests (#6320)
• The bundled Traefik helm chart has been updated to v18.0.0 (#6353)
• Mark v1.25.3+k3s1 as stable (#6338)
• The embedded helm controller has been bumped to v0.13.0 (#6294)
• Fixed an issue that would prevent the deploy controller from handling manifests that include resource types that are no longer supported by the apiserver. (#6295)
• Replace fedora-coreos with fedora 36 for install tests (#6315)
• Convert containerd config.toml.tmpl Linux template to v2 syntax (#6267)
• Add test for node-external-ip config parameter (#6359)
• Use debugger-friendly compile settings if DEBUG is set (#6147)
• update e2e tests (#6354)
• Remove unused vagrant development scripts (#6395)
• The bundled Traefik has been updated to v2.9.4 / helm chart v18.3.0 (#6397)
• None (#6371)
• Fix incorrect defer usage (#6296)
• Add snapshot restore e2e test (#6396)
• Fix sonobouy tests on v1.25 (#6399)
• Bump packaged component versions
• The packaged traefik helm chart has been bumped to v19.0.0, enabling ingressClass support by default.
• The packaged local-path-provisioner has been bumped to v0.0.23
• The packaged coredns has been bumped to v1.9.4 (#6408)
• log kube-router version when starting netpol controller (#6405)
• Add Kairos to ADOPTERS (#6417)
• Update Flannel to 0.20.1 (#6388)
• Avoid wrong config for flannel-external-ip and add warning if unencrypted backend (#6403)
• Fix test-mods to allow for pinning version from k8s.io (#6413)
• Fix for metrics-server in the multi-cloud cluster env (#6386)
• K3s now indicates specifically which cluster-level configuration flags are out of sync when critical configuration differs between server nodes. (#6409)
• Convert test output to JSON format (#6410)
• Pull traefik helm chart directly from GH (#6468)
• Nightly test fix (#6475)
• Update to v1.25.4 (#6477)
• Remove stuff which belongs in the windows executor implementation (#6492)
• The packaged traefik helm chart has been bumped to 19.0.4 (#6494)
• Move traefik chart repo again (#6508)

Release v1.25.3+k3s1

This release updates Kubernetes to v1.25.3, and fixes a number of issues.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.2+k3s1:

• E2E: Groundwork for PR runs (#6131)
• Fix flannel for deployments of nodes which do not belong to the same network and connect using their public IP (#6180)
• Mark v1.24.6+k3s1 as stable (#6193)
• Add cluster reset test (#6161)
• The embedded metrics-server version has been bumped to v0.6.1 (#6151)
• The ServiceLB (klipper-lb) service controller is now integrated into the K3s stub cloud controller manager. (#6181)
• Events recorded to the cluster by embedded controllers are now properly formatted in the service logs. (#6203)
• Fix error dialing backend errors in apiserver network proxy (#6216)
  • Fixed an issue with the apiserver network proxy that caused kubectl exec to occasionally fail with error dialing backend: EOF
  • Fixed an issue with the apiserver network proxy that caused kubectl exec and kubectl logs to fail when a custom kubelet port was used, and the custom port was blocked by firewall or security group rules.
• Fix the typo in the test (#6183)
• Use setup-go action to cache dependencies (#6220)
• Add journalctl logs to E2E tests (#6224)
• The embedded Traefik version has been bumped to v2.9.1 / chart 12.0.0 (#6223)
• Fix flakey etcd test (#6232)
• Replace deprecated ioutil package (#6230)
• Fix dualStack test (#6245)
• Add ServiceAccount for svclb pods (#6253)
• Update to v1.25.3-k3s1 (#6269)
• Return ProviderID in URI format (#6284)
• Corrected CCM RBAC to allow for removal of legacy service finalizer during upgrades. (#6306)
• Added a new —flannel-external-ip flag. (#6321)
  • When enabled, Flannel traffic will now use the nodes external IPs, instead of internal.
  • This is meant for use with distributed clusters that are not all on the same local network.

Release v1.25.2+k3s1

This release updates Kubernetes to v1.25.2, and fixes a number of issues.

For more details on what’s new, see the Kubernetes release notes.

Changes since v1.25.0+k3s1:

• Add k3s v1.25 to the release channel (#6129)
• Restore original INSTALL_K3S_SKIP_DOWNLOAD behavior (#6130)
• Add K3S Release Documentation (#6135)
• Update to v1.25.1 (#6140)
• Update to v1.25.2-k3s1 (#6168)

Release v1.25.0+k3s1

This release is K3S’s first in the v1.25 line. This release updates Kubernetes to v1.25.0.

Before upgrading from earlier releases, be sure to read the Kubernetes Urgent Upgrade Notes.

Important Note: Kubernetes v1.25 removes the beta PodSecurityPolicy admission plugin. Please follow the upstream documentation to migrate from PSP if using the built-in PodSecurity Admission Plugin, prior to upgrading to v1.25.0+k3s1.

Changes since v1.24.4+k3s1:

• Update Kubernetes to v1.25.0 (#6040)
• Remove --containerd flag from windows kubelet args (#6028)
• E2E: Add support for CentOS 7 and Rocky 8 (#6015)
• Convert install tests to run PR build of k3s (#6003)
• CI: update Fedora 34 -> 35 (#5996)
• Fix dualStack test and change ipv6 network prefix (#6023)
• Fix e2e tests (#6018)
• Update README.md (#6048)
• Remove wireguard interfaces when deleting the cluster (#6055)
• Add validation check to confirm correct golang version for Kubernetes (#6050)
• Expand startup integration test (#6030)
• Update go.mod version to 1.19 (#6049)
• Usage of --cluster-secret, --no-deploy, and --no-flannel is no longer supported. Attempts to use these flags will cause fatal errors. See the docs for their replacement. (#6069)
• Update Flannel version to fix older iptables version issue. (#6090)
• The bundled version of runc has been bumped to v1.1.4 (#6071)
• The embedded containerd version has been bumped to v1.6.8-k3s1 (#6078)
• Fix deprecation message (#6112)
• Added warning message for flannel backend additional options deprecation (#6111)