A few nodes of the pulsar storage layer known as bookkeeper crashed and propagate the pulsar cluster with them. We are restoring the bookkeeper cluster and then we will help the cluster pulsar to recover.
EDIT 19:15 UTC : We have deployed a patch to fix the bookkeeper cluster, we have deployed the new configuration and we are rolled out the cluster. The pulsar cluster should be available.
EDIT 20:20 UTC : Some nodes of the bookkeeper cluster are under memory pressure, we are investigating the issue.
EDIT 21:20 UTC: We found the issue and are deploying the patch.
EDIT 21:50 UTC: Situation is back to normal.