All systems are operational.
The issue has been resolved.
Due to retiring an old storage cluster node, we found a misconfiguration on our XMPP server that used the IP of that node instead of the proper gateway address.
The administrator is notified and the problem should fixed soon.