Update the Collaboration Manager

Update the Management Portal

Update to a newer version

1. Create a backup of the backend database before updating.
2. Check the Release Notes for any breaking changes or special instructions.

3. Either download the helm chart from the GitHub Releases or check-out the revision of the version you'd like to update to:

   git checkout ${VERSION:?}
   

4. Upgrade the Helm Chart, either by using the downloaded chart:

   helm upgrade \
       --dependency-update \
       -f values.yaml \
       ${RELEASE_NAME:?} \
       collab-manager-${VERSION:?}$.tgz
   

   or using the local Git repository:

   helm upgrade \
       --dependency-update \
       -f values.yaml \
       ${RELEASE_NAME:?} \
       ./helm
   

5. Wait until all containers are up & running. If a container can't be started anymore, check the logs of the container for more information or roll-back the update.

Roll-back to a previous version

1. Find out if there were any database migrations between the current version and the version you'd like to roll-back to.

   You can do so with the GitHub Compare feature. Select the current version and the version you'd like to roll-back and check for any changes in the backend/capellacollab/alembic directory.

   If there are no changes, skip the next steps and run the update as usual.

2. Scale down the backend deployment to 0:

   kubectl scale deployment ${RELEASE_NAME:?}-backend --replicas=0
   

3. Restore the backup of the backend database to the state it was before the update as described in the restore documentation.

   Danger

   This will discard any changes made to the database after the update.

4. Run the update as usual and use the version you'd like to roll-back to as target version.

Update the Postgres database

Warning

Major PostgreSQL database updates can only be done with downtime. Plan it accordingly and communicate the downtime to your users in advance.

If database.backend.deploy is set to True in your values.yaml, follow these instructions:

1. Create a backup of the backend database.
2. Make sure that the backup contains real data and is not empty.
3. Change the database.backend.internal.postgresVersion to the desired postgres version in the values.yaml.

4. Scale down the PostgreSQL deployment to 0:

   kubectl scale deployment ${RELEASE_NAME:?}-backend-postgres --replicas=0
   

5. Delete the persistent volume claim (PVC) of the PostgreSQL instance:

   kubectl delete pvc ${RELEASE_NAME:?}-volume-backend-postgres
   

6. Run a helm upgrade as described in the "Update to a newer version" section.

7. Restore the database backup as described in the restore documentation.

If database.backend.deploy is set to False in your values.yaml, follow these instructions:

1. Create a backup of the backend database.
2. Upgrade the external PostgreSQL instance to the desired version. Please refer to the PostgreSQL documentation for instructions on how to do this.
3. If the PostgreSQL connection string has changed, change the database.backend.external.uri field in the values.yaml to the new connection URI. Then, run a helm upgrade as described in the "Update to a newer version" section.