--- title: "Full move for Kubernetes universes" url: "https://docs.yugabyte.com/stable/yugabyte-platform/scale-deployments/kubernetes-full-move/" --- # Full move for Kubernetes universes Modify storage class, volume count, and volume size on running YugabyteDB Anywhere Kubernetes universes. Change volume attributes on operator and non-operator Kubernetes universes Full move for Kubernetes universes lets you modify storage attributes such as volume count, storage class, and volume size on existing universes. Decreasing volume size is supported and is carried out through a full move (persistent volumes cannot be shrunk in place). Version requirement Full move for Kubernetes universes requires YugabyteDB v2026.1.0.0 or later on the universe. Universes on earlier database versions cannot use this feature. Use this page based on how your universe is managed: - [Non-operator universes](#non-operator-universes "Non-operator universes"): Universes created and managed through the YugabyteDB Anywhere UI or API using [Helm charts](/stable/yugabyte-platform/scale-deployments/edit-helm-overrides/ "Helm charts"). - [Operator universes](#operator-universes "Operator universes"): Universes managed with the [YugabyteDB Kubernetes Operator](/stable/yugabyte-platform/anywhere-automation/yb-kubernetes-operator/ "YugabyteDB Kubernetes Operator"). Change volumes by updating the `YBUniverse` CRD. If you [imported a Helm-managed universe](/stable/yugabyte-platform/anywhere-automation/yb-kubernetes-operator/#import-universe "imported a Helm-managed universe") to the Operator, use the Operator workflow going forward. ## Non-operator universes Full move applies to universes that YugabyteDB Anywhere manages via Helm charts. For information on creating these universes and setting initial volume configuration, refer to [Create universes](/stable/yugabyte-platform/create-deployments/create-universes-wizard/ "Create universes") and [Configure a Kubernetes provider](/stable/yugabyte-platform/configure-yugabyte-platform/kubernetes/ "Configure a Kubernetes provider"). ### Modify volume attributes - New UI - Classic UI [New UI](/stable/yugabyte-platform/yba-overview/#new-experience) To change storage class, volume count, or volume size on a running universe, navigate to the universe and do the following: 1. Click **Settings > Hardware**, and under **Cluster Instance**, or **TServer Instance** or **Master Server Instance** if the universe has dedicated masters, click **Edit**. 2. Edit the volume and storage fields as needed. 3. To edit per-AZ storage overrides, under **Storage Overrides** click **Add Storage Overrides** or **Edit**. [Classic UI](/stable/yugabyte-platform/yba-overview/#new-experience) To change storage class, volume count, or volume size on a running universe, navigate to the universe and do the following: 1. Click **Actions > Edit Universe**. 2. Edit the volume fields for TServer and Master as needed. ![Edit instance configuration](/images/ee/fm-instance-config.png "Edit instance configuration") 3. To edit per-AZ storage overrides, use **Edit storage overrides** in the UI. ![Storage overrides](/images/ee/edit-storage-overrides.png "Storage overrides") Monitor the **Edit Kubernetes Universe** task on the universe [Tasks](/stable/yugabyte-platform/manage-deployments/retry-failed-task/ "Tasks") page. To change other Helm chart settings (resources, labels, and so on) without changing storage class or volume count, use [Edit Kubernetes overrides](/stable/yugabyte-platform/scale-deployments/edit-helm-overrides/ "Edit Kubernetes overrides") instead. ### How volume attributes are determined *Existing universes* are migrated so that the volume attributes in use are stored as per-availability zone (AZ) overrides and applied on each Helm operation. *New universes* maintain backward compatibility. When deciding per-AZ volume attributes, YugabyteDB Anywhere considers all sources, including [provider per-AZ storage class](/stable/yugabyte-platform/configure-yugabyte-platform/kubernetes/#configure-region-and-zones "provider per-AZ storage class") and [provider overrides](/stable/yugabyte-platform/configure-yugabyte-platform/kubernetes/#overrides "provider overrides"). For storage class recommendations, refer to [Hardware requirements for pods](/stable/yugabyte-platform/prepare/server-nodes-hardware/ "Hardware requirements for pods"). When you add a new AZ, volume attributes for that zone are populated from the same sources the first time. After that, persisted attributes are used. For each AZ, the effective volume attributes are the merge of that AZ's overrides and the base fields `deviceInfo` (TServer) and `masterDeviceInfo` (Master), with AZ overrides taking precedence. The UI may show `standard` for the storage class in base volume attributes when the field was not populated. The values in use come from AZ overrides merged with the base values. Modifying base volume settings when AZ overrides exist Because AZ overrides take complete precedence, changes to base volume fields when modifying universe settings only affect AZs that *do not* have their own overrides configured. If an AZ already has volume overrides defined, the base change has no effect on that zone. If all AZs in the universe have volume overrides configured, modifying base volume values has no effect. That is, no AZ uses the base settings. To change volume attributes in that case, edit the storage overrides to update the per-AZ overrides directly. ### Example Non-operator universes store per-AZ volume overrides under `userIntent.userIntentOverrides.azOverrides`. Each AZ is identified by a UUID. Under each AZ, `perProcess` can define `TSERVER` and `MASTER` `deviceInfo` overrides (for example, `storageClass`, `numVolumes`, `volumeSize`). The following illustrates `userIntent` with `userIntentOverrides.azOverrides` after a volume edit. It is stored configuration, not a file you upload. ```json "userIntent": { "numNodes": 3, "ybSoftwareVersion": "2026.1.0.0-b1", "accessKeyCode": "yugabyte-default", "deviceInfo": { "volumeSize": 50, "numVolumes": 1, "storageClass": "standard" }, "masterDeviceInfo": { "volumeSize": 50, "numVolumes": 1, "storageClass": "standard" }, "userIntentOverrides": { "azOverrides": { "ecabfc93-6559-4c53-8d6b-aa66a08e8263": { "perProcess": { "TSERVER": { "deviceInfo": { "storageClass": "yb-standard" } }, "MASTER": { "deviceInfo": { "storageClass": "yb-standard" } } } }, "c4143bba-96ce-4a4c-9b4c-9f9d4fd6860f": { "perProcess": { "TSERVER": { "deviceInfo": { "storageClass": "vk-standard" } }, "MASTER": { "deviceInfo": { "storageClass": "yb-standard" } } } } } } } ``` ## Operator universes To use full move on universes managed by the [YugabyteDB Kubernetes Operator](/stable/yugabyte-platform/anywhere-automation/yb-kubernetes-operator/ "YugabyteDB Kubernetes Operator"), configure the `tserverVolume` and `masterVolume` fields on the `YBUniverse` CRD (`volumeSize`, `numVolumes`, `storageClass`). Each field supports a `perAZ` section for AZ-specific values. Changing any of these attributes, including decreasing `volumeSize` triggers a full move. For creating operator-managed universes, refer to [Create a universe](/stable/yugabyte-platform/anywhere-automation/yb-kubernetes-operator/#create-a-universe "Create a universe"). Per-AZ storage class on the provider is configured through [Create a provider](/stable/yugabyte-platform/anywhere-automation/yb-kubernetes-operator/#create-a-provider "Create a provider") (`kubernetesStorageClass`). If you migrated a Helm-managed universe to the Operator using [Import universe](/stable/yugabyte-platform/anywhere-automation/yb-kubernetes-operator/#import-universe "Import universe"), edit volume settings on the `YBUniverse` CRD; the UI blocks most edit actions on imported universes. When `perAZ` overrides are present, volume attributes for an AZ are the merge of that AZ's `perAZ` values and the base `tserverVolume` or `masterVolume` fields, with `perAZ` taking precedence. If `perAZ` values are absent for an AZ, the base volume attribute values apply entirely. When `perAZ` overrides are not used, the provider storage class applies. If the provider has no storage class, YugabyteDB Anywhere falls back to the base `tserverVolume` and `masterVolume` fields. After you use the `perAZ` section once, you cannot remove it from the CRD. To use only the base values for all AZs again, set `perAZ` to an empty object (`perAZ: {}`). For field definitions and schema details, run `kubectl describe crd ybuniverse`. ### Example The following `YBUniverse` spec uses `tserverVolume` and `masterVolume` with `perAZ` overrides for `us-west1-a`: ```yaml apiVersion: operator.yugabyte.io/v1alpha1 kind: YBUniverse metadata: name: universe-1 spec: numNodes: 3 replicationFactor: 3 enableYSQL: true enableNodeToNodeEncrypt: true enableClientToNodeEncrypt: true ybSoftwareVersion: 2.31.0.0-b30 enableYSQLAuth: false enableYCQL: true enableYCQLAuth: false enableIPV6: false enableLoadBalancer: false tserverVolume: numVolumes: 1 volumeSize: 80 perAZ: us-west1-a: volumeSize: 80 numVolumes: 1 storageClass: vk-standard masterVolume: numVolumes: 1 volumeSize: 60 perAZ: us-west1-a: volumeSize: 60 numVolumes: 1 storageClass: vk-standard ``` Apply changes with `kubectl apply` to the `YBUniverse` resource. The Operator reconciles the universe and runs the full move workflow. Monitor tasks in YugabyteDB Anywhere on the universe [Tasks](/stable/yugabyte-platform/manage-deployments/retry-failed-task/ "Tasks") page. ## Batch pod moves Full move can move pods in batches instead of all at once. Control batching with the **Number of nodes to move in a given batch during full move** universe runtime configuration option (config key `yb.task.full_move.roll_batch_size`). Refer to [Manage runtime configuration settings](/stable/yugabyte-platform/administer-yugabyte-platform/manage-runtime-config/ "Manage runtime configuration settings"). The following table describes how each setting for `yb.task.full_move.roll_batch_size` affects how pods are moved during a full move: | Value | Behavior | |---------------|------------------------------------------------------------------------------| | `0` (default) | All pods that must move are moved in a single operation. | | Non-zero | Each availability zone moves pods in parallel, in batches of the given size. |