hi i m suffering from an issue that we went to production without any limit for osd_max_pg_log_entries and osd_min_pg_log_entries (nautilus version) the problem is that the memory of the osd got very very high with many objects and big number of PG's i saw in similar case recommendation for this but i do not understand based on which parameters like : osd device size ? block.db size ? osd_min_pg_log_entries=500 osd_max_pg_log_entries=500 bluefs_buffered_io=false https://tracker.ceph.com/issues/53729 can someone share more light , what are the parameters for decided those values ? osd_min_pg_log_entries=500 osd_max_pg_log_entries=500 Side note: we understand that Once we have a big number of Pg log entries , the limit does snot update on the fly (running system) and not evenm on boot And using manual log trim procedure is required - am I right ? _______________________________________________ Dev mailing list -- dev@xxxxxxx To unsubscribe send an email to dev-leave@xxxxxxx
