Manage Clusters

Palette supports several day 2 operations to manage the end-to-end lifecycle of the Kubernetes clusters launched through Palette. It also provides several capabilities across new and imported clusters to keep your clusters secure, compliant, up-to-date, and perform ongoing management operations like backup/restore. Additionally, you can have visibility into the workloads running inside your cluster and cluster costs.

The following sections describe these capabilities in detail:

  • Reconfigure - Scale your clusters up/down by adding/reducing the number of nodes in a node pool and adding additional worker pools. Resize nodes in a node pool by modifying the node specs (CPU, Memory, or Instance Type for public clouds). Add additional fault domains such as availability zones to a node pool.
  • Updates - Add/remove add-ons; Upgrade versions of the core packs (Kubernetes, CSI, CNI); Upgrade versions of add-on layers (Monitoring, Security, etc.)
  • Cluster Health Alerts - Palette monitors the health of all workload clusters and raises an alert when the cluster goes to an unhealthy state. Besides displaying the alert on the UI console, Palette provides the ability to have these alerts pushed out to a variety of channels. Users can set up simple email alerts to receive an email when health status of their cluster changes.
  • Compliance Scans - Perform continuous compliance checks to ensure your clusters are secure, conformant, and compliant.
  • OS Patching - Automatically apply the most recent security patches to cluster nodes to stay up to date with latest OS patches.
  • Backup and Restore - Backup your cluster configurations as well as any persistent volumes used by your applications regularly. Choose critical namespaces you would like to back up. Restore as required to new or existing clusters.
  • Cost Visibility - Get visibility into the estimated cloud cost for the cluster based on cluster node configuration. Get additional insights into per namespace cost (Usage Cost) calculated based on the amount of resources consumed within the namespace.
  • Workload Visibility - Palette provides visibility into the resources running inside workload clusters. These resources are displayed on the cluster details page.
  • Node Labels and Taints - You can constrain a Pod to only run on a particular set of Node(s). There are several ways to do this and the recommended approaches such as, nodeSelector, node affinity, etc. all use label selectors to facilitate the selection. Taints allow a node to repel a set of pods for appropriate pod allocation to node pools.
  • RBAC and NS Support - RoleBindings and ClusterRoleBindings are Role-Based Access Control (RBAC) concepts that allow granular control over cluster-wide resources as well as namespaced resources. Palette provides the ability to specify these bindings to configure granular RBAC rules. Palette also can define new namespaces for the cluster and manage (remove, assign quota, assign role bindings, etc.) them.
  • Add-on Pack Status and Monitoring - Packs associated with the cluster you are monitoring are displayed with information on the status and the installation progress of the Add-on Packs. The pack statuses appear as gray indicating as an initial onboarding and before deployment status, green representing a successful installation, blue as processing the installation, and red as a failed install and requires attention.
  • Web based kubectl - Palette leverages Kubectl through an in-built command line interface for the users to communicate with their workload clusters.
  • Platform Management - Palette supports the Cluster(s) Management feature to exclude a cluster or a group of clusters from getting upgraded when Palette is upgraded.
  • NOC UI - Palette provides Intuitive UI-based location monitoring for the clusters running at multiple locations. For public cloud clusters Palette displays the Region set during the cluster creation process and displays the location on the UI Map. For private cloud clusters the user can set the location through the Palette UI. The user can monitor the location details of all the clusters running under a specific scope.
Palette supports new and existing clusters across public and private clouds. Furthermore, in public clouds, Palette supports managed Kubernetes cluster services such as EKS, AKS, etc. The capabilities described above may not be available for all cluster types or may work slightly different in some cluster types. Relevant sections will highlight these exceptions.