Upgrade a cluster in Kubernetes

This page shows how to upgrade a CockroachDB cluster that is .

Overview

Types of upgrades

• Major-version upgrades: A , such as from v24.2 to v24.3, may include new features, updates to cluster setting defaults, and backward-incompatible changes. Performing a major-version upgrade requires an additional step to finalize the upgrade.

  As of 2024, there are four major versions per year, where every second release is an . For CockroachDB Advanced, Innovation releases include all of the latest features but have shorter and can be skipped.

  For CockroachDB Standard and Basic, only Regular releases are supported. Major version upgrades are applied automatically to Standard and Basic clusters as they become available, and Standard clusters have the option to turn off automatic updates.

  As of 2024, every second major version is an . Innovation releases offer shorter support windows and can be skipped.

• Patch upgrades: A moves a cluster from one patch release to another within a major version, such as from v24.2.3 to v24.2.4. Patch upgrades do not introduce backward-incompatible changes.

  A major version of CockroachDB has two phases of patch releases: a series of testing releases (beta, alpha, and RC releases) followed by a series of production releases (vX.Y.0, vX.Y.1, and so on). A major version’s first production release (the .0 release) is also known as its GA release.

  For all CockroachDB Cloud clusters, production patch releases for a major version are automatically applied until the cluster is upgraded to a new major version. On an Advanced cluster, you have the option to defer patch updates for 30, 60, or 90 days if you have a configured.

To learn more about CockroachDB major versions and patches, refer to .

Upgrade differences across Cloud plans

CockroachDB Cloud plan	Major version upgrades	Innovation releases
Basic	Automatic	Unavailable
Standard	Automatic (default) or customer-initiated	Unavailable
Advanced	Customer-initiated	Optional

For CockroachDB Basic clusters and CockroachDB Standard clusters that are set to upgrade automatically, major version upgrades are finalized immediately and cannot be rolled back. Basic and Standard clusters are upgraded to the .1 patch release of a major version, not the initial .0 release.

For all Cloud plans:

• All major versions that are Regular releases (as opposed to Innovation releases) are required upgrades before proceeding to a newer major version. To maintain support, an upgrade to a supported version must occur prior to the of the current version.
• Patch version upgrades occur automatically.

Compatible versions

A cluster may always be upgraded to the next major Regular release once it is made available in CockroachDB Cloudrelease. Every second major version is an Innovation release that can be deployed or skipped on CockroachDB Advanced clusters:

• If your cluster is running a major version that is a Regular release, it can be upgraded to either:

  • the next major version (an Innovation release)
  • the release that follows the next major version (the next Regular release, once it is available, skipping the Innovation release).

• If a cluster is running a major version that is labeled an Innovation release, it can be upgraded only to the next Regular release.

On Kubernetes, the upgrade is a staged update in which each pod’s container image for CockroachDB is updated in a rolling fashion. The cluster remains available during the upgrade.

Select the cluster’s deployment method to continue.

Before you begin

If you , substitute kubectl with oc in the following commands.

All kubectl steps should be performed in the . By default, this is cockroach-operator-system.

Before beginning a major-version or patch upgrade:

1. Verify the overall health of your cluster using the :
   • Under Node Status, make sure all nodes that should be live are listed as such. If any nodes are unexpectedly listed as SUSPECT or DEAD, identify why the nodes are offline and either restart them or them before beginning your upgrade. If there are DEAD and non-decommissioned nodes in your cluster, the upgrade cannot be finalized. If any node is not fully decommissioned, try the following:
     1. First, reissue the . The second command typically succeeds within a few minutes.
     2. If the second decommission command does not succeed, and then decommission it again. Before continuing the upgrade, the node must be marked as decommissioned.
   • Under Replication Status, make sure there are 0 under-replicated and unavailable ranges. Otherwise, performing a rolling upgrade increases the risk that ranges will lose a majority of their replicas and cause cluster unavailability. Therefore, it’s important to identify and resolve the cause of range under-replication and/or unavailability before beginning your upgrade.
   • In the Node List, make sure all nodes are on the same version. Upgrade them to the cluster’s current version before continuing. If any nodes are behind, this also indicates that the previous major-version upgrade may not be finalized.
   • In the Metrics dashboards, make sure , , and capacity are within acceptable values for each node. Nodes must be able to tolerate some increase in case the new version uses more resources for your workload. If any of these metrics is above healthy limits, consider to your cluster before beginning your upgrade.
2. Make sure your cluster is behind a , or your clients are configured to talk to multiple nodes. If your application communicates with only a single node, stopping that node to upgrade its CockroachDB binary will cause your application to fail.
3. By default, the uses a compaction concurrency of 3. If you have sufficient IOPS and CPU headroom, you can consider increasing this setting via the COCKROACH_COMPACTION_CONCURRENCY environment variable. This may help to reshape the LSM more quickly in scenarios; and it can lead to increased overall performance for some workloads. Cockroach Labs strongly recommends testing your workload against non-default values of this setting.

are automated backups of CockroachDB Cloud clusters that are stored by Cockroach Labs in cloud storage. By default, Cockroach Labs takes and retains managed backups in all Cloud clusters.

When upgrading to a major release, you can optionally of your cluster to your own cloud storage, as an extra layer of protection in case the upgrade leads to issues.

CockroachDB is designed with high fault tolerance. However, taking regular backups of your data is an operational best practice for planning. Refer to .

1. Review the , as well as the release notes for any skipped major version. Pay careful attention to the sections for backward-incompatible changes, deprecations, changes to default cluster settings, and features that are not available until the upgrade is finalized.
2. Optionally disable auto-finalization to preserve the ability to roll back a major-version upgrade instead of finalizing it. If auto-finalization is disabled, a major-version upgrade is not complete until it is finalized.

Ensure you have a valid license key

To perform major version upgrades, you must have .

Patch version upgrades can be performed without a valid license key, with the following limitations:

• The cluster will run without limitations for a specified grace period. During that time, alerts are displayed that the cluster needs a valid license key. For more information, refer to the .
• The cluster is at the end of the grace period if no before then.

If you have an Enterprise Free or Enterprise Trial license, you must enable telemetry using the cluster setting, as shown below in order to finalize a major version upgrade:

SET CLUSTER SETTING diagnostics.reporting.enabled = true;

If a cluster with an Enterprise Free or Enterprise Trial license is upgraded across patch versions and does not meet telemetry requirements:

• The cluster will run without limitations for a 7-day grace period. During that time, alerts are displayed that the cluster needs to send telemetry.
• The cluster is if telemetry is not received before the end of the grace period.

For more information, refer to the .

If you want to stay on the previous version, you can roll back the upgrade before finalization.

Perform a patch upgrade

To upgrade from one patch release to another within the same major version, perform the following steps on one node at a time:

1. Change the container image in the custom resource:

   image:
     name: cockroachdb/cockroach:
   

2. Apply the new settings to the cluster:

   kubectl apply -f example.yaml
   

   The Operator will perform the staged update.

3. To check the status of the rolling upgrade, run kubectl get pods.

4. Verify that all pods have been upgraded:

   $ kubectl get pods \
   -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.spec.containers[0].image}{"\n"}'
   

   You can also check the CockroachDB version of each node in the .

5. Add a partition to the update strategy defined in the StatefulSet. Only the pods numbered greater than or equal to the partition value will be updated. For a cluster with 3 pods (e.g., cockroachdb-0, cockroachdb-1, cockroachdb-2) the partition value should be 2:

   $ kubectl patch statefulset cockroachdb \
   -p='{"spec":{"updateStrategy":{"type":"RollingUpdate","rollingUpdate":{"partition":2}}}}'
   
   statefulset.apps/cockroachdb patched
   

6. Change the container image in the StatefulSet:

   $ kubectl patch statefulset cockroachdb \
   --type='json' \
   -p='[{"op": "replace", "path": "/spec/template/spec/containers/0/image", "value":"cockroachdb/cockroach:"}]'
   
   statefulset.apps/cockroachdb patched
   

7. To check the status of the rolling upgrade, run kubectl get pods.

8. Verify that all pods have been upgraded:

   $ kubectl get pods \
   -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.spec.containers[0].image}{"\n"}'
   

   You can also check the CockroachDB version of each node in the .

9. Add a partition to the update strategy defined in the StatefulSet. Only the pods numbered greater than or equal to the partition value will be updated. For a cluster with 3 pods (e.g., cockroachdb-0, cockroachdb-1, cockroachdb-2) the partition value should be 2:

   $ helm upgrade \
   my-release \
   cockroachdb/cockroachdb \
   --set statefulset.updateStrategy.rollingUpdate.partition=2
   

10. Connect to the cluster using the SQL shell:

    $ kubectl exec -it cockroachdb-client-secure \
    -- ./cockroach sql \
    --certs-dir=/cockroach-certs \
    --host=my-release-cockroachdb-public
    

11. Remove the cluster initialization job from when the cluster was created:

    $ kubectl delete job my-release-cockroachdb-init
    

12. Change the container image in the StatefulSet:

    $ helm upgrade \
    my-release \
    cockroachdb/cockroachdb \
    --set image.tag= \
    --reuse-values
    

13. To check the status of the rolling upgrade, run kubectl get pods.

14. Verify that all pods have been upgraded:

    $ kubectl get pods \
    -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.spec.containers[0].image}{"\n"}'
    

    You can also check the CockroachDB version of each node in the .

Roll back a patch upgrade

To roll back a patch upgrade, repeat the steps in Perform a patch upgrade, but configure the container image for the pods to the previous major version.

Perform a major-version upgrade

To perform a major upgrade:

1. Change the container image image in the custom resource:

   image:
     name: cockroachdb/cockroach:
   

2. Apply the new settings to the cluster:

   kubectl apply -f example.yaml
   

   The Operator will perform the staged update.

3. To check the status of the rolling upgrade, run kubectl get pods.

4. Verify that all pods have been upgraded:

   $ kubectl get pods \
   -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.spec.containers[0].image}{"\n"}'
   

   You can also check the CockroachDB version of each node in the .

5. Before beginning a major-version upgrade, the Operator disables auto-finalization by setting the cluster setting cluster.preserve_downgrade_option to the cluster’s current major version. Before finalizing an upgrade, follow your organization’s testing procedures to decide whether to finalize or roll back the upgrade. After finalization begins, you can no longer roll back to the cluster’s previous major version.

7. Add a partition to the update strategy defined in the StatefulSet. Only the pods numbered greater than or equal to the partition value will be updated. For a cluster with 3 pods (e.g., cockroachdb-0, cockroachdb-1, cockroachdb-2) the partition value should be 2:

   $ kubectl patch statefulset cockroachdb \
   -p='{"spec":{"updateStrategy":{"type":"RollingUpdate","rollingUpdate":{"partition":2}}}}'
   
   statefulset.apps/cockroachdb patched
   

8. Change the container image in the StatefulSet:

   $ kubectl patch statefulset cockroachdb \
   --type='json' \
   -p='[{"op": "replace", "path": "/spec/template/spec/containers/0/image", "value":"cockroachdb/cockroach:"}]'
   
   statefulset.apps/cockroachdb patched
   

9. To check the status of the rolling upgrade, run kubectl get pods.

10. After the pod has been restarted with the new image, start the CockroachDB :

    $ kubectl exec -it cockroachdb-client-secure \-- ./cockroach sql \
    --certs-dir=/cockroach-certs \
    --host=cockroachdb-public
    

11. Run the following SQL query to verify that the number of underreplicated ranges is zero:

    SELECT sum((metrics->>'ranges.underreplicated')::DECIMAL)::INT AS ranges_underreplicated FROM crdb_internal.kv_store_status;
    
      ranges_underreplicated
    --------------------------
                           0
    (1 row)
    

    This indicates that it is safe to proceed to the next pod.

12. Exit the SQL shell:

    > \q
    

13. Decrement the partition value by 1 to allow the next pod in the cluster to update:

    $ kubectl patch statefulset cockroachdb \
    -p='{"spec":{"updateStrategy":{"type":"RollingUpdate","rollingUpdate":{"partition":1}}}}'
    
    statefulset.apps/cockroachdb patched
    

14. Repeat steps 4-8 until all pods have been restarted and are running the new image (the final partition value should be 0).

15. Check the image of each pod to confirm that all have been upgraded:

    $ kubectl get pods \
    -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.spec.containers[0].image}{"\n"}'
    
    cockroachdb-0   cockroachdb/cockroach:
    cockroachdb-1   cockroachdb/cockroach:
    cockroachdb-2   cockroachdb/cockroach:
    ...
    

    You can also check the CockroachDB version of each node in the .

16. If auto-finalization is disabled, the upgrade is not complete until you finalize the upgrade.

17. Add a partition to the update strategy defined in the StatefulSet. Only the pods numbered greater than or equal to the partition value will be updated. For a cluster with 3 pods (e.g., cockroachdb-0, cockroachdb-1, cockroachdb-2) the partition value should be 2:

    $ helm upgrade \
    my-release \
    cockroachdb/cockroachdb \
    --set statefulset.updateStrategy.rollingUpdate.partition=2
    

18. Connect to the cluster using the SQL shell:

    $ kubectl exec -it cockroachdb-client-secure \
    -- ./cockroach sql \
    --certs-dir=/cockroach-certs \
    --host=my-release-cockroachdb-public
    

19. Remove the cluster initialization job from when the cluster was created:

    $ kubectl delete job my-release-cockroachdb-init
    

20. Change the container image in the StatefulSet:

    $ helm upgrade \
    my-release \
    cockroachdb/cockroachdb \
    --set image.tag= \
    --reuse-values
    
    NAME                                READY     STATUS              RESTARTS   AGE
    my-release-cockroachdb-0            1/1       Running             0          2m
    my-release-cockroachdb-1            1/1       Running             0          3m
    my-release-cockroachdb-2            0/1       ContainerCreating   0          25s
    my-release-cockroachdb-init-nwjkh   0/1       ContainerCreating   0          6s
    ...
    

Ignore the pod for cluster initialization. It is re-created as a byproduct of the StatefulSet configuration but does not impact your existing cluster.

1. After the pod has been restarted with the new image, start the CockroachDB :

   $ kubectl exec -it cockroachdb-client-secure \
   -- ./cockroach sql \
   --certs-dir=/cockroach-certs \
   --host=my-release-cockroachdb-public
   
   $ kubectl run cockroachdb -it \
   --image=cockroachdb/cockroach \
   --rm \
   --restart=Never \
   -- sql \
   --insecure \
   --host=my-release-cockroachdb-public
   

2. Run the following SQL query to verify that the number of underreplicated ranges is zero:

   SELECT sum((metrics->>'ranges.underreplicated')::DECIMAL)::INT AS ranges_underreplicated FROM crdb_internal.kv_store_status;
   
     ranges_underreplicated
   --------------------------
                          0
   (1 row)
   

   This indicates that it is safe to proceed to the next pod.

3. Exit the SQL shell:

   > \q
   

4. Decrement the partition value by 1 to allow the next pod in the cluster to update:

   $ helm upgrade \
   my-release \
   cockroachdb/cockroachdb \
   --set statefulset.updateStrategy.rollingUpdate.partition=1 \
   

5. Repeat steps 4-8 until all pods have been restarted and are running the new image (the final partition value should be 0).

6. Check the image of each pod to confirm that all have been upgraded:

   $ kubectl get pods \
   -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.spec.containers[0].image}{"\n"}'
   
   my-release-cockroachdb-0    cockroachdb/cockroach:
   my-release-cockroachdb-1    cockroachdb/cockroach:
   my-release-cockroachdb-2    cockroachdb/cockroach:
   ...
   

   You can also check the CockroachDB version of each node in the .

7. If auto-finalization is disabled, the upgrade is not complete until you finalize the upgrade.

Finalize a major-version upgrade manually

To finalize a major-version upgrade:

1. Connect to the cluster using the SQL shell:

   $ kubectl exec -it cockroachdb-client-secure \
   -- ./cockroach sql \
   --certs-dir=/cockroach-certs \
   --host=cockroachdb-public
   

2. Run the following command:

   > RESET CLUSTER SETTING cluster.preserve_downgrade_option;
   

   A series of migration jobs runs to enable certain types of features and changes in the new major version that cannot be rolled back. These include changes to system schemas, indexes, and descriptors, and enabling certain types of improvements and new features. Until the upgrade is finalized, these features and functions will not be available and the command SHOW CLUSTER SETTING version will return the previous version`.

   You can monitor the process of the migration in the DB Console . Migration jobs have names in the format -{migration-id}. If a migration job fails or stalls, Cockroach Labs can use the migration ID to help diagnose and troubleshoot the problem. Each major version has different migration jobs with different IDs.

   The amount of time required for finalization depends on the amount of data in the cluster, because finalization runs various internal maintenance and migration tasks. During this time, the cluster will experience a small amount of additional load.

   When all migration jobs have completed, the upgrade is complete.

3. To confirm that finalization has completed, check the cluster version:

   > SHOW CLUSTER SETTING version;
   

   If the cluster continues to report that it is on the previous version, finalization has not completed. If auto-finalization is enabled but finalization has not completed, check for the existence of where decommission has stalled. In most cases, issuing the decommission command again resolves the issue. If you have trouble upgrading, contact Support.

Roll back a major-version upgrade

To roll back to the previous major version before an upgrade is finalized:

1. Change the container image in the custom resource to use the previous major version:

   image:
     name: cockroachdb/cockroach:v25.1
   

2. Apply the new settings to the cluster:

   kubectl apply -f example.yaml
   

   The Operator will perform the staged rollback.

3. To check the status of the rollback, run kubectl get pods.

4. Verify that all pods have been rolled back:

   $ kubectl get pods \
   -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.spec.containers[0].image}{"\n"}'
   

Rollbacks do not require finalization.

Disable auto-finalization

For clusters managed by the Public operator, auto-finalization is disabled and cannot be enabled. A major version upgrade is not complete until it is manually finalized. The Public operator does not support the cluster.auto_upgrade.enabled.

By default, auto-finalization is enabled, and a major-version upgrade is finalized when all nodes have rejoined the cluster using the new cockroach binary. This means that by default, a major-version upgrade cannot be rolled back. Instead, you must .

To disable auto-finalization:

1. Connect to the cluster using the SQL shell:

   $ kubectl exec -it cockroachdb-client-secure \
   -- ./cockroach sql \
   --certs-dir=/cockroach-certs \
   --host=cockroachdb-public
   

2. Set the cluster.auto_upgrade.enabled to false.

Now, to complete a major-version upgrade, you must manually finalize it or roll it back.

Previously, to disable automatic finalization and preserve the ability to roll back a major-version upgrade, it was required to set the cluster setting cluster.preserve_downgrade_option to the cluster’s current major version before beginning the major-version upgrade, and then to unset the setting to finalize the upgrade.

We now recommend managing a cluster’s finalization policy using the cluster setting cluster.auto_upgrade.enabled, which was introduced in v23.2. The setting does not need to be modified after it is initially set.

Either of these settings prevents automatic finalization.

By default, auto-finalization is enabled, and a major-version upgrade is finalized when all nodes have rejoined the cluster using the new cockroach binary. This means that by default, a major-version upgrade cannot be rolled back. Instead, you must .

To disable auto-finalization:

1. Connect to the cluster using the SQL shell:

   $ kubectl exec -it cockroachdb-client-secure \
   -- ./cockroach sql \
   --certs-dir=/cockroach-certs \
   --host=cockroachdb-public
   

2. Set the cluster.preserve_downgrade_option to the cluster’s current major version. The Helm chart does not yet support the cluster.auto_upgrade.enabled.

Now, to complete a major-version upgrade, you must manually finalize it or roll it back.

By default, auto-finalization is enabled, and a major-version upgrade is finalized when all nodes have rejoined the cluster using the new cockroach binary. This means that by default, a major-version upgrade cannot be rolled back. Instead, you must .

To disable auto-finalization:

1. Connect to the cluster using the SQL shell:

   kubectl exec -it cockroachdb-client-secure \
   -- ./cockroach sql \
   --certs-dir=/cockroach-certs \
   --host=cockroachdb-public
   

2. Set the cluster.auto_upgrade.enabled to false.

Now, to complete a major-version upgrade, you must manually finalize it or roll it back.

Previously, to disable automatic finalization and preserve the ability to roll back a major-version upgrade, it was required to set the cluster setting cluster.preserve_downgrade_option to the cluster’s current major version before beginning the major-version upgrade, and then to unset the setting to finalize the upgrade.

We now recommend managing a cluster’s finalization policy using the cluster setting cluster.auto_upgrade.enabled, which was introduced in v23.2. The setting does not need to be modified after it is initially set.

Either of these settings prevents automatic finalization.

By default, auto-finalization is enabled, and a major-version upgrade is finalized when all nodes have rejoined the cluster using the new cockroach binary. This means that by default, a major-version upgrade cannot be rolled back. Instead, you must .

To disable auto-finalization:

1. Connect to the cluster using the SQL shell:

   cockroach sql
   

2. Set the cluster.auto_upgrade.enabled to false.

Now, to complete a major-version upgrade, you must manually finalize it or roll it back.

Previously, to disable automatic finalization and preserve the ability to roll back a major-version upgrade, it was required to set the cluster setting cluster.preserve_downgrade_option to the cluster’s current major version before beginning the major-version upgrade, and then to unset the setting to finalize the upgrade.

We now recommend managing a cluster’s finalization policy using the cluster setting cluster.auto_upgrade.enabled, which was introduced in v23.2. The setting does not need to be modified after it is initially set.

Either of these settings prevents automatic finalization.

Troubleshooting

After the upgrade has finalized (whether manually or automatically), it is no longer possible to roll back the upgrade. If you are experiencing problems, we recommend that you for assistance.

In the event of catastrophic failure or corruption, it may be necessary to from a backup to a new cluster running the previous major version.

See also