Changelog since 2.2.0-1.13.3
- Kubernetes v1.13.4
- Docker v18.09.3
- Calico v3.5.2
- Support for Mesos pre-reserved roles for etcd, control-plane, public-node and private-node, and placement rules for etcd.
- Modify how etcd placement constraints are defined, there is now a separate
etcd.placement option. For backwards compatibility, if it’s left empty the value from
kubernetes.control_plane_placement will be used.
- Fix a bug where sometimes Kubernetes workloads running on public agents would not have access to Kubernetes workloads running on private agents.
- Fix a bug where using
dcos kubernetes cluster kubeconfig resulted in an improperly encoded
certificate-authority-data in the generated
Changelog since 2.1.1-1.12.5
- Fix the
etcd snapshotting instructions.
- Add section
Mesos Roles to
Advanced Installation page.
- Kubernetes v1.13.3
- dcos-commons v0.55.2
- CoreDNS v1.3.1
- Enable CSI features required for CSI integration.
- Automate the task replacement when a DC/OS agent is decommissioned.
- Allow changing automated DC/OS proxy configuration into Kubernetes cluster tasks.
- Fix a bug where providing
cluster backup and
cluster restore commands did not actually work.
- Fix a bug affecting clusters in which the Kubernetes service CIDR or Calico network CIDR overlapped with Docker’s default bridge network by disabling the bridge.
- Add a Storage page documenting Container Storage Interface (CSI).
Changelog since 2.1.0-1.12.3
- dcos-commons v0.55.0
- Kubernetes v1.12.5
- Docker v18.09.1
- Kubernetes Dashboard v1.10.1
- Enable local-dns-dispatcher in control plane tasks.
- Fix a bug that might cause pods that have resource limits crash on RHEL based systems. The issue is related to Linux kmem accounting turned-on by default by runc. We now turn-off kmem accounting on RHEL-based systems, and on these systems alone. No user intervention is needed, however all of Kubernetes cluster tasks will be replaced, which may cause some downtime.
Changelog since 2.0.1-1.12.2
- dcos-commons v0.54.3
- Kubernetes v1.12.3
- CoreDNS v1.2.6
- Calico v3.2.4
etcd. When set, etcd will check all incoming peer requests from the cluster for valid client certificates signed by the supplied CA.
- Enable the selection of the desired region where to deploy the Kubernetes cluster.
- Add the new flag
--force in the
cluster update command to force the update of the cluster configuration.
- Support relative paths in
cluster kubeconfig command, e.g.
- Move the validation of the service configuration to the Mesosphere Kubernetes Engine.
cluster backup and
cluster restore commands. The AWS session token can now be used as part of the AWS credentials.
- Increase the number of retries an etcd task will perform during installation to resolve its own DNS name. This should prevent etcd tasks from getting stuck in a retry loop on larger clusters.
- Fix a bug that might cause segfault when running
dcos kubernetes cluster kubeconfig.
- Documentation section on how to upgrade the
Changelog since 2.0.0-1.12.1
- dcos-commons v0.54.2.
- Kubernetes v1.12.2
- Fix a bug affecting use of private Docker registries.
Changelog since 1.x
- Kubernetes v1.12.1 and other components’ version changes.
- Enable high density deployments of multiple Kubernetes clusters on DC/OS.
kube-scheduler tasks with static pods in a new task
coredns tasks with static pods, and rename
- Replace cluster DNS
kube-dns deployment with
coredns and prevent co-location of these pods.
- Add the
priorityClassName field to critical system pods.
- The Kubernetes Dashboard is now secured using HTTPS and will now show the login view when accessed.
- Use a dedicated RBAC role for the
- Add options to enable Calico’s Typha.
- Public Kubernetes nodes now reserve ports
443 of the underlying public DC/OS agent to help prevent issues with port binding, and to making them available for Ingress.
- Installation and package options upgrades are now faster.
- Scaling up a cluster is now performed in parallel and therefore faster. Scaling down a cluster is still performed serially to ensure workload stability while decommissioning Kubernetes nodes.
- Fix a bug that might cause
kube-node-public tasks to freeze in the
STARTED state, causing installations or upgrades to stop indefinitely.
- Fix a bug that could forever fail to run public Kubernetes node tasks.
- Fix a bug affecting node decommission that could cause Kubernetes apps temporary downtime.