Hello Experts,
Trying to optimize the replication performances, we increased concurrency_backup setting. this have helped, but there is a step that is taking a lot of time : vbr Finalizing object replication.
What is happening during this step and how can it be optimized?
Below an extract from logs (attached the full log).
[dbadmin@lvrtkj10 ~]$ cat /appli/VRTK/log/vbr/vbr_2019-01-23-202531.log 2019-01-23 20:25:31 localhost vbr VBR log initialized. [...] 2019-01-23 20:25:46 localhost vbr Snapshotting the source database. 2019-01-23 20:26:15 localhost vsql Warning from vsql: WARNING 7814: Snapshotting at epoch 4004408. Current epoch is: 4004410, More recent data will not be included. 2019-01-23 20:33:13 localhost vbr Snapshot complete. [...] 2019-01-23 20:33:16 localhost vbr Copying catalog snapshot from source to destination. 2019-01-23 20:34:03 localhost vbr Snapshot version: v9.1.0-2 (releases/VER_9_1_RELEASE_BUILD_0_2_20180524) 2019-01-23 20:34:03 localhost vbr Preparing destination database for replicating objects. 2019-01-23 20:35:46 localhost vbr Prep complete, start syncing files. 2019-01-23 20:36:31 localhost vbr Approximate bytes to copy: 623678573064 of 20435179761567 total. 2019-01-23 20:46:10 localhost vbr Complete syncing files, removing snapshot from the source database. 2019-01-23 20:46:14 localhost vbr Finalizing object replication. 2019-01-23 22:08:08 localhost vbr Object replication complete! 2019-01-23 22:08:08 localhost vsql /opt/vertica/bin/vsql -q -t -X -dVRTK_UNI_KJ -p5433 -Udbadmin -h192.168.0.1
Many thanks.
mkheir
