k3s-io/k3s v1.31.0+k3s1

on GitHub

This release is K3S's first in the v1.31 line. This release updates Kubernetes to v1.31.0.

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

Changes since v1.30.4+k3s1:

• Move test-compat docker test to GHA (#10414)
• Check for bad token permissions when install via PR (#10387)
• Bump k3s-root to v0.14.0 (#10466)
  • The k3s bundled userspace has been bumped to a release based on buildroot 2024.02.3, addressing several CVEs in busybox and coreutils.
• Fix INSTALL_K3S_PR support (#10472)
• Add data-dir to uninstall and killall scripts (#10473)
• Bump github.com/hashicorp/go-retryablehttp from 0.7.4 to 0.7.7 (#10400)
• Bump golang:alpine image version (#10359)
• Bump Local Path Provisioner version (#10394)
• Ensure remotedialer kubelet connections use kubelet bind address (#10480)
  • Fixed an issue where setting the --bind-address flag to a non-loopback or wildcard address would prevent kubectl logs from working properly.
• Bump Trivy version (#10339)
• Add etcd s3 config secret implementation (#10340)
  • A proxy can now be configured for use when uploading etcd snapshots to a s3-compatible storage service. This overrides any proxy settings passed via environment variables.
  • Credentials and endpoint configuration for storing etcd snapshots on a s3-compatible storage service can now be read from a Secret, instead of passing them via the CLI or config file. See https://github.com/k3s-io/k3s/blob/master/docs/adrs/etcd-s3-secret.md for more information.
• For E2E upgrade test, automatically determine the channel to use (#10461)
• Bump kine to v0.11.11 (#10494)
• Fix loadbalancer reentrant rlock (#10511)
  • Fixed an issue that could cause the agent loadbalancer to deadlock when the currently in-use server goes down.
• Don't use server value from config file for etcd-snapshot commands (#10514)
  • The --server and --token flags for the k3s etcd-snapshot command have been renamed to --etcd-server and --etcd-token, to avoid unintentionally running snapshot management commands against a remote node when the cluster join address or token are present in a config file.
• Use pagination when listing large numbers of resources (#10527)
• Fix multiple issues with servicelb (#10552)
  • Fixed issue that caused ServiceLB to fail to create a daemonset for services with long names
  • Fixed issue that caused ServiceLB pods to crashloop on nodes with ipv6 disabled at the kernel level
• Enhance E2E Hardened option (#10558)
• Allow Pprof and Superisor metrics in standalone mode (#10576)
• Use higher QPS for secrets reencryption (#10571)
• Fix issues loading data-dir value from env vars or dropin config files (#10591)
• Remove deprecated use of wait. functions (#10546)
• Wire lasso metrics up to metrics endpoint (#10528)
• Update stable channel to v1.30.3+k3s1 (#10647)
• Bump docker/docker to v25.0.6 (#10642)
• Add a change for killall to not unmount server and agent directory (#10403)
• Allow edge case OS rpm installs (#10680)
• Bump containerd to v1.7.20 (#10659)
• Update to newer OS images for install testing (#10681)
• Bump helm-controller to v0.16.3 to drop Helm v2 support (#10628)
• Add toleration support to ServiceLB DaemonSet (#10687)
  • • New Feature: Users can now define Kubernetes tolerations for ServiceLB DaemonSet directly in the svccontroller.k3s.cattle.io/tolerations annotation on services.
• Fix: Add $SUDO prefix to transactional-update commands in install script (#10531)
• Update to v1.30.3-k3s1 and Go 1.22.5 (#10707)
• Fix caching name for e2e vagrant box (#10695)
• Fix k3s-killall.sh support for custom data dir (#10709)
• Adding MariaDB to README.md (#10717)
• Bump Trivy version (#10670)
• V1.31.0-k3s1 (#10715)
• Update kubernetes to v1.31.0-k3s3 (#10780)

Embedded Component Versions

Helpful Links

As always, we welcome and appreciate feedback from our community of users. Please feel free to:

• Open issues here
• Join our Slack channel
• Check out our documentation for guidance on how to get started or to dive deep into K3s.
• Read how you can contribute here