I am not able to run replication yet, waiting for a port to be open. But why
should I see SizeOfLogQueue and Replication Lag increase? I have not
enabled replication on any table. Just added a peer. No tables:columns are
set to replicate yet.
Interesting that TimeStampsOfLastAppliedOp shows the date when this cluster
was setup to replicate to the same target (last time logs were applied was
indeed on Jan 23). But since Jan 23 I completely rebuilt the target cluster
(blew away all data and built it from scratch but this time pointing to
another zookeeper cluster). I removed the old peer on this source and added
the new one (this one pointing now to a new zookeeper).
Like somehow it kept some metadata about the old peer. Not sure how's that
possible when I removed it. I think that removes all the metadata for it.
Showing a sample here:
hbase(main):001:0> status 'replication'
version 1.2.0
32 live servers
hb01dn13.pod01.xyz.com:
SOURCE: PeerID=3, AgeOfLastShippedOp=0, SizeOfLogQueue=6,
TimeStampsOfLastShippedOp=Thu Jan 01 00:00:00 UTC 1970, Replication
Lag=1485897034081
SINK : AgeOfLastAppliedOp=0, TimeStampsOfLastAppliedOp=Mon Jan 23
18:40:46 UTC 2017
hb01dn11.pod01.xyz.com:
SOURCE: PeerID=3, AgeOfLastShippedOp=0, SizeOfLogQueue=7,
TimeStampsOfLastShippedOp=Thu Jan 01 00:00:00 UTC 1970, Replication
Lag=1485897034622
SINK : AgeOfLastAppliedOp=0, TimeStampsOfLastAppliedOp=Tue Jan 31
15:44:50 UTC 2017
hb01dn26.pod01.xyz.com:
SOURCE: PeerID=1, AgeOfLastShippedOp=0, SizeOfLogQueue=25,
TimeStampsOfLastShippedOp=Thu Jan 01 00:00:00 UTC 1970, Replication
Lag=1485897033131
SINK : AgeOfLastAppliedOp=0, TimeStampsOfLastAppliedOp=Mon Jan 23
18:40:46 UTC 2017
--
View this message in context:
http://apache-hbase.679495.n3.nabble.com/replication-concepts-enabling-peer-vs-enabling-table-replication-tp4085975p4085985.html
Sent from the HBase User mailing list archive at Nabble.com.
