Posts for: #k8s

K3s Releases Latest Version: v1.29.1+k3s2

K3s Releases Latest Version: v1.29.1+k3s2

K3s, the lightweight, highly available, certified Kubernetes distribution, has released version v1.29.1+k3s2. This release is designed for production workloads in unattended, resource-constrained, remote locations, or inside IoT appliances. K3s is packaged as a single <70MB binary, reducing the dependencies and steps needed to install, run, and auto-update a production Kubernetes cluster.

The update to Kubernetes v1.29.1 brings several fixes and improvements. Some of the changes since v1.29.0+k3s1 include:

  • Bump Sonobuoy version
  • Bump actions/setup-go from 4 to 5
  • Update stable channel to v1.28.5+k3s1 and add v1.29 channel
  • Added support for env *_PROXY variables for agent loadbalancer
  • Add a retry around updating a secrets-encrypt node annotations
  • Silence SELinux warning on INSTALL_K3S_SKIP_SELINUX_RPM
  • Add ServiceLB support for PodHostIPs FeatureGate
  • Redirect error stream to null when checking nm-cloud systemd unit
  • Dockerfile.dapper: set $HOME properly
  • Add system-agent-installer-k3s step to GA release instructions
  • Fix install script checksum
  • and many more…

For a full list of changes, you can check the Kubernetes release notes.

This release also includes updates to various embedded components, such as Kubernetes v1.29.1, Kine v0.11.0, SQLite 3.42.0, etcd v3.5.9-k3s1, containerd v1.7.11-k3s2, runc v1.1.12-k3s1, Flannel v0.24.0, metrics-server v0.6.3, Traefik v2.10.5, CoreDNS v1.10.1, helm-controller v0.15.8, and local-path-provisioner v0.0.24.

K3s Unveils New Version: v1.29.0+k3s1

K3s Unveils New Version: v1.29.0+k3s1

K3s has released version v1.29.0+k3s1, an update to its lightweight, highly available Kubernetes distribution. This release includes several important changes and updates, including an upgrade to Kubernetes v1.29.0. However, before upgrading, users are advised to read the Urgent Upgrade Notes from Kubernetes.

There are two important changes to note in this release. The first is the removal of the experimental rotate-keys subcommand, due to changes in Kubernetes upstream for KMSv2. This subcommand may be added back in future releases. The second change is the removal of the multi-cluster-cidr flag, as support for this alpha feature has been completely removed from Kubernetes upstream.

Other notable changes in this release include fixing an overlapping address range, updating the stable channel to November 2023, adding runtime classes for wasm/nvidia/crun, and bumping containerd/runc to v1.7.10-k3s1/v1.1.10. Additionally, there are updates to containerd, removal of GA feature-gates, improvements to code coverage, and an update to flannel.

Embedded component versions in this release include Kubernetes v1.29.0, Kine v0.11.0, SQLite 3.42.0, Etcd v3.5.9-k3s1, Containerd v1.7.11-k3s2, Runc v1.1.10, Flannel v0.24.0, Metrics-server v0.6.3, Traefik v2.10.5, CoreDNS v1.10.1, Helm-controller v0.15.4, and Local-path-provisioner v0.0.24.

Flux Version 2.2.0 Released

Flux Version 2.2.0 Released

Flux CD has released version 2.2.0, a tool for keeping Kubernetes clusters in sync with sources of configuration and automating updates to configuration when there is new code to deploy. This feature release brings several updates and improvements to provide users with the best experience.

The Flux CLI and controllers have been updated to support Kustomize v5.3.0 and Kubernetes v1.28.4. This ensures compatibility with the latest versions and improves performance and stability.

One of the major updates in this release is the significant overhaul of the Flux helm-controller’s reconciliation model. This addresses persistent issues such as the automatic recovery of releases stuck in a pending state. It also improves the observability of the release status and introduces the ability to enable drift detection on a per-object basis. More details on the helm-controller improvements can be found in the Announcing Flux 2.2 GA blog post.

The Flux CLI now allows users to force or reset the reconciliation state of a HelmRelease v2beta2 object using the flux reconcile hr --force and flux reconcile hr --reset commands. This gives users more control over the deployment process.

The Flux CLI also comes with support for bootstrapping Gitea repositories, providing users with more options for managing their configuration sources. It also adds guardrails to flux install and flux bootstrap to protect users from destructive operations. Additionally, the flux version and flux check commands now print the Flux distribution version deployed on the cluster, making it easier for users to keep track of their Flux installation.

The alerting capabilities of Flux have been extended with support for NATS and Bitbucket Server & Data Center. This allows users to use these additional notification providers for alerting and monitoring purposes.

Starting with this release, Flux minor versions are benchmarked to measure the Mean Time To Production (MTTP), providing users with valuable insights into the performance of Flux.

The release is compatible with Kubernetes versions 1.26, 1.27, and 1.28.

The API changes in this release include the promotion of the HelmRelease kind from v2beta1 to v2beta2. The v2beta2 API is backwards compatible with v2beta1, and the v2beta1 API is deprecated and will be removed in a future release. Several new fields have been added, including drift detection and correction on a per-release basis, selective running of Helm tests, and a history of metadata from Helm releases. The Alert and Provider kinds have also been promoted from v1beta2 to v1beta3, with the removal of the .status field. The Bucket API now has a new field for server-side filtering of files, and the OCIRepository and HelmChart APIs have new fields for verifying OIDC identity. The HelmRepository and ImageRepository APIs have a new boolean field for connecting to non-TLS HTTP container registries.

To upgrade to Flux v2.2.0, users can either rerun the bootstrap process or use the Flux GitHub Action. The APIs can be upgraded by deploying the new CRDs and controllers and changing the manifests in Git. It is advised not to delay this procedure as the deprecated versions will be removed after 6 months.

The release also includes new documentation for the HelmRelease v2beta2 specification and a guide on enabling in-memory Kustomize builds.

Overall, Flux v2.2.0 brings several updates and improvements to provide users with a better experience in managing their Kubernetes clusters and configuration sources. Users are encouraged to upgrade to this version for the best experience.

K3s Unveils New Version: v1.28.4+k3s1

K3s, a lightweight and highly available Kubernetes distribution, has released version v1.28.4+k3s1. This certified Kubernetes distribution is specifically designed for production workloads in resource-constrained and unattended environments, such as remote locations or IoT appliances. The new release updates Kubernetes to v1.28.4 and includes several fixes for various issues.

Some of the changes and improvements in this release include:

  • Update channels latest to v1.27.7+k3s2
  • Add etcd status condition for easy monitoring of etcd status from each node
  • Automatic discovery of WebAssembly runtimes
  • Improved dualStack log
  • Optimized Dockerfile for simplified installation and runtime
  • Addition of timezone info in the Docker image, enabling the use of spec.timeZone in CronJobs
  • Bumped kine to v0.11.0, resolving issues with postgres and NATS, improving watch channel performance, and enhancing compatibility
  • QoS-class resource configuration for containerd
  • Addition of agent flag disable-apiserver-lb to disable load balance proxy
  • Various bug fixes and improvements

For a full list of changes, please refer to the Kubernetes release notes.

The embedded component versions in this release are as follows:

  • Kubernetes v1.28.4
  • Kine v0.11.0
  • SQLite 3.42.0
  • Etcd v3.5.9-k3s1
  • Containerd v1.7.7-k3s1
  • Runc v1.1.8
  • Flannel v0.22.2
  • Metrics-server v0.6.3
  • Traefik v2.10.5
  • CoreDNS v1.10.1
  • Helm-controller v0.15.4
  • Local-path-provisioner v0.0.24

For more information and resources on K3s, you can visit the official documentation, join the Slack channel, or contribute to the project on GitHub.

K3s Unveils Latest Version v1.28.3+k3s2

K3s, a lightweight and highly available certified Kubernetes distribution, has released version v1.28.3+k3s2. This version is designed for production workloads in resource-constrained and remote locations, as well as inside IoT appliances. K3s comes as a single binary that is less than 70MB in size, making it easy to install, run, and auto-update a production Kubernetes cluster.

The latest release updates Kubernetes to version v1.28.3 and addresses several issues. Some of the changes since v1.28.3+k3s1 include:

  • Restoration of selinux context systemd unit file
  • Update of channel to v1.27.7+k3s1
  • Bump of Sonobuoy version
  • Bump of Trivy version
  • Fix for accessing outer scope .SystemdCgroup, which resolves issues with starting with nvidia-container-runtime
  • Upgrade of traefik chart to v25.0.0
  • Update of traefik to fix registry value
  • Improvement to not use iptables-save/iptables-restore if it will corrupt rules

The components and versions included in this release are as follows:

  • Kubernetes v1.28.3
  • Kine v0.10.3
  • SQLite 3.42.0
  • Etcd v3.5.9-k3s1
  • Containerd v1.7.7-k3s1
  • Runc v1.1.8
  • Flannel v0.22.2
  • Metrics-server v0.6.3
  • Traefik v2.10.5
  • CoreDNS v1.10.1
  • Helm-controller v0.15.4
  • Local-path-provisioner v0.0.24

For more information on the release and its features, refer to the Kubernetes release notes.

Overall, this new release of K3s brings important updates and fixes to enhance the performance and reliability of Kubernetes clusters in production environments.

Source: K3s

Longhorn Unveils Latest Update: Longhorn v1.4.4 Release

Longhorn has released version v1.4.4, a distributed block storage system for Kubernetes. This release includes various enhancements, improvements, bug fixes, and stability and resilience updates. Some of the notable improvements include the addition of disk status Prometheus metrics, improved log levels for resource update failures, and support for both NFS hard and soft with custom timeo and retrans options for RWX volumes. Bugs related to volume synchronization, attaching/detaching loops, and volume mounting have also been addressed. This release is aimed at providing a more stable and reliable storage solution for Kubernetes environments. For more information, you can visit the Longhorn v1.4.4 release page.