Have you looked at the log file to try and understand what's taking so long? For example, what happens here: Thursday 01 October 2020 12:02:54 +0530 (0:00:04.528) 0:09:57.974 ****** Thursday 01 October 2020 12:15:42 +0530 (0:12:47.271) 0:22:45.246 ****** - seems to be RUNNING HANDLER [ceph-handler : restart ceph osds daemon(s)] And here? Thursday 01 October 2020 12:26:50 +0530 (0:00:04.525) 0:33:53.382 ****** Thursday 01 October 2020 12:39:40 +0530 (0:12:50.003) 0:46:43.385 ****** Which seems to be another restart. Now that we know this is slow, can you look at why?