Datalore 2024.1 Help

Server lifecycle events

This topic describes Datalore server lifecycle events, such as restart or shutdown.

As Datalore's pod is controlled by a StatefulSet, there are two ways to restart Datalore - by deleting the pod manually, or by doing the same by invoking the scaling mechanism.

kubectl scale statefulset/datalore --replicas 0 kubectl wait pod/datalore-0 --for=delete --timeout=600s

Once the second command is completed, it's completely safe to interact with Datalore's database or storage (for example, for backup purposes).

To get it back online, the StatefulSet should be scaled back up to 1:

kubectl scale statefulset/datalore --replicas 1 kubectl wait pod/datalore-0 --for=condition=Ready --timeout=600

Alternatively, a delete action could be invoked:

kubectl delete pod/datalore-0 kubectl wait pod/datalore-0 --for=delete --timeout=600s

As the Datalore deployment is controlled by Docker Compose, a native Compose mechanism should be used for the lifecycle events.

# start server docker compose up # gracefully stop and start back docker compose restart # gracefully stop docker compose stop
Last modified: 30 January 2024