Hi all, I have a two ceph 13.2.6 clusters in multisite setup on HDD disks with ~466.0 M objects and rather low usage: 63 MiB/s rd, 1.5 MiB/s wr, 978 op/s rd, 308 op/s wr. In each cluster there are two dedicated rgws for repliaction (setted as zone endpoints, other rgws have "rgw run sync thread = false")
Replication lag is about 15sek: Master at Thu Aug 29 09:30:21 CEST 2019: metadata sync no sync (zone is master) data sync source: master_zone syncing full sync: 0/128 shards incremental sync: 128/128 shards data is behind on 35 shards behind shards: [8,10,11,29,34,37,45,46,52,54,56,57,58,59,60,62,70,77,78,80,81,82,88,89,90,94,96,97,105,109,112,119,122,125,127] oldest incremental change not applied: 2019-08-29 09:30:03.0.537216s 70 shards are recovering recovering shards: [0,1,2,5,6,7,9,10,11,12,13,14,16,18,20,21,22,23,25,28,29,34,35,36,37,39,43,48,49,52,53,54,55,56,57,58,59,60,61,63,65,67,68,69,70,72,75,76,83,84,86,91,92,97,99,101,104,105,109,110,111,112,115,116,117,119,120,121,122,126] Slave at Thu Aug 29 09:30:22 CEST 2019: metadata sync syncing full sync: 0/64 shards incremental sync: 64/64 shards metadata is caught up with master data sync source: slave_zone syncing full sync: 0/128 shards incremental sync: 128/128 shards data is behind on 24 shards behind shards: [11,12,15,18,20,35,51,57,59,60,67,82,83,84,86,89,93,97,105,108,120,122,125,127] oldest incremental change not applied: 2019-08-29 09:30:11.0.755569s 64 shards are recovering recovering shards: [0,1,2,3,6,7,8,9,10,11,13,14,15,16,20,21,22,23,25,27,28,29,35,36,37,38,39,43,46,48,49,52,56,59,60,61,62,63,65,67,68,69,70,76,79,83,84,85,88,90,91,97,100,104,105,109,110,111,113,117,118,120,122,123] Is there anything to speed-up replication? Should I enable "rgw run sync thread" on all rgws not just zone endpoints? Best Regards, Tom
_______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com