[ceph-users] ceph-mgr high CPU utilization

I'm wondering if anyone still sees issues with ceph-mgr using CPU and being unresponsive even in recent Nautilus releases.  We upgraded our largest cluster from Mimic to Nautilus (14.2.8) recently - it has about 3500 OSDs.  Now ceph-mgr is constantly at 100-200% CPU (1-2 cores), and becomes unresponsive after a few minutes.  The finisher-Mgr queue length grows (I've seen it at over 100k) - similar symptoms as seen with earlier Nautilus releases by many.  This is what it looks like after an hour of running:     "finisher-Mgr": {         "queue_len": 66078,         "complete_latency": {             "avgcount": 21,             "sum": 2098.408767721,             "avgtime": 99.924227034         }     }, We have a pretty vanilla manager config, only the balancer is enabled in upmap mode.  Here are the enabled modules:     "always_on_modules": [         "balancer",         "crash",         "devicehealth",         "orchestrator_cli",         "progress",         "rbd_support",         "status",         "volumes"     ],     "enabled_modules": [         "restful"     ], Any ideas or outstanding issues in this area? Andras