Hi! osd_heartbeat_interval indicates interval (6 seconds) between peer pings, if peer does not reply within osd_heartbeat_grace(20 seconds), osd will report peer osd failure to mon, and then mon to mark down failure osd. So, the client request will be blocked within 20 seconds, 20 seconds is too long. If we adjust osd_heartbeat_grace and osd_heartbeat_interval as follow: osd_heartbeat_grace = 7 osd_heartbeat_interval = 3 When the peer pings failure, the client request will be stuck for 7 seconds. Is there any risk in adjusting the osd_heartbeat_grace and osd_heartbeat_interval, or other better best practices. Best regard _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx
 |