From 27514d58573ce1f844af4ea60afa72c7f58f1cd7 Mon Sep 17 00:00:00 2001 From: Jasper Ras Date: Thu, 21 Aug 2025 22:09:29 +0200 Subject: vault backup: 2025-08-21 22:09:29 --- Cluster issues openstack.md | 11 +++++++++++ 1 file changed, 11 insertions(+) create mode 100644 Cluster issues openstack.md (limited to 'Cluster issues openstack.md') diff --git a/Cluster issues openstack.md b/Cluster issues openstack.md new file mode 100644 index 0000000..100e086 --- /dev/null +++ b/Cluster issues openstack.md @@ -0,0 +1,11 @@ +[[Resetting cluster members]] + +--- + +[Mirantis](https://docs.mirantis.com/mcp/q4-18/mcp-operations-guide/openstack-operations/rabbitmq-ops.html) +> We recommend that you do not restart a RabbitMQ node on a production environment by executing **systemctl restart rabbitmq-server** since a cluster can become inoperative. + +Check `rabbitmqctl list_queues messages consumers name` if the dashboard shows a lot of ready messages it might be that there are queues where messages are being sent to but have no consumers. They expire after 30 minutes so wait for that first and then see. + +A lot of unacknowledged messages on a specific node: `rabbitmq-upgrade drain` and then `rabbitmq-upgrade revive` seems to help, but in some cases the symptom moves to another node. + -- cgit v1.2.3