From 400abea6616693286aa4385db9da1b232547e56b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Vedran=20Ivankovi=C4=87?= <33936733+Veki301@users.noreply.github.com> Date: Thu, 10 Jul 2025 13:22:09 +0200 Subject: [PATCH] update migrate-team-features documentation --- .../administrate/migrate-team-features.md | 48 ++++++++++--------- 1 file changed, 25 insertions(+), 23 deletions(-) diff --git a/src/how-to/administrate/migrate-team-features.md b/src/how-to/administrate/migrate-team-features.md index 97a7f23..40a9430 100644 --- a/src/how-to/administrate/migrate-team-features.md +++ b/src/how-to/administrate/migrate-team-features.md @@ -1,39 +1,39 @@ # Migrate to new team features -This document describes how to migrate your team settings if the release is older than [5.12.0](https://docs.wire.com/latest/changelog/changelog.html#2025-03-06-chart-release-5120) in the air-gapped environment. If you do not perform this migration after deploying a newer release, your team settings will be in an inconsistent state, until this is run. - -We added a [migrate-features](https://github.com/wireapp/helm-charts/tree/main/charts/migrate-features) helm chart to facilitate the migration process. +Release of [wire-server-5.12](https://docs.wire.com/latest/changelog/changelog.html#2025-03-06-chart-release-5120) introduces a new data storage format for team features. To migrate to the new format, we have added a [tool](https://github.com/wireapp/helm-charts/tree/main/charts/migrate-features) called `migrate-features` to our deployment bundle, which will have to be run, after upgrading to a `wire-server` release past 5.12. Migration has to be run only once. +If you do not perform this migration after deploying to a release past `5.12`, your team settings will be in an inconsistent state. ## Configuration - The values of the chart `migrate-features/values.yaml` already contains the necessary configuration options. - - ```yaml - job: - name: migrate-features - restartPolicy: Never - backoffLimit: 4 - cassandraHost: cassandra-external # Replace with your Cassandra service name `kubectl get svc -n default` - cassandraPort: 9042 - galleyKeyspace: galley - ``` +The values of the chart `migrate-features/values.yaml` already contains the necessary configuration options. + +```yaml +job: + name: migrate-features + restartPolicy: Never + backoffLimit: 4 + cassandraHost: cassandra-external # Replace with your Cassandra service name `kubectl get svc -n default` + cassandraPort: 9042 + galleyKeyspace: galley +``` + Most configuration values can remain as they are, but ensure the `cassandraHost` value is correct by running: ```sh kubectl get svc -n default` # assuming the services are in the default namespace ``` -And the service name for the cassandra and replace the value of `cassandraHost` if it differs. + +If your cassandra installation differs from the standard `wire-server-deploy` setup, a hostname or an IP address of one of the cassandra nodes can be supplied here. ## Run the migration job When this chart is installed, it sets up a `migrate-feature` job in the `default` namespace: - ```sh - helm install default ./charts/migrate-features - ``` - -Before running the helm command, make sure the `migrate-features` chart is available in the charts directory, if not please download the chart or copy it from our latest artifacts. +```sh +helm install default ./charts/migrate-features +``` +Before running the helm command, make sure the `migrate-features` chart is available in the charts directory, if not please [download](https://github.com/wireapp/helm-charts/tree/main/charts/migrate-features) the chart or copy it from one of our latest artifact. ## Troubleshoot and monitoring @@ -43,9 +43,9 @@ After the helm installation check if the job is running kubectl get jobs -n default ``` -When the job is complete, you will see a `migrate-features` job with a `COMPLETION` status of `1/1`. This may take some time. If the job does not complete, check the pod logs or Kubernetes events. +When the job is complete, you will see a `migrate-features` job with a `COMPLETIONS` status of `1/1`. This may take some time. If the job does not complete, check the pod logs or Kubernetes events. -To check the k8s events run: +To check the pod events run: ```sh kubectl describe job migrate-features -n default @@ -55,4 +55,6 @@ To check the logs run: ```sh kubectl logs job/migrate-features -n default -``` \ No newline at end of file +``` + +While the migration tool is running, team features are going to operate in read-only mode for the team that is currently being migrated. After migration, the new storage is going to be used. No other action should be required on the part of instance operators besides running the migration tool.