Google Groups

svc-extractor-filters and replicator.filter.replicate.do doesn't seem to work as expected


Dima Ilchenko Jan 23, 2012 9:19 PM
Posted in group: Tungsten Replicator Discuss
Hello,

trying to set up tungsten replicator in direct mode to takeover mysql
replication (single master and slave) which only replicates few
databases from master. According to documentation in wiki, that's
doable with replicator.filter.replicate.do in tungsten-replicator
2.0.5. This is a sharded database setup, so on master there are
databases named db001,db002...db064.
Here's a command used to install tungsten-replicator (version
2.0.5-483):

tools/tungsten-installer --direct -a --service-name=parallel --native-
slave-takeover \
--master-host=master-db.ourdomain --master-port=3306 --master-
user=tungsten --master-password=password \
--slave-host=127.0.0.1 --slave-port=13325 --slave-user=root \
--home-directory=/opt/tungsten  \
--property=replicator.store.parallel-queue.maxOfflineInterval=5  \
--property=replicator.auto_enable=false \
--svc-extractor-filters=replicate \
--property=replicator.filter.replicate.do=db001,db002 \
--svc-parallelization-type=disk --buffer-size=100 --channels=10  \
--start-and-report

Right after this, status shows it's in OFFLINE:NORMAL state. As soon
as '$TUNGSTEN_BIN/trepctl -service parallel online' command is issued,
status becomes OFFLINE:ERROR, and the exception message says it failed
to execute statement which inserts data into, for example, db010 which
is clearly not included into replicator.filter.replicate.do list.
At the same time, ' $TUNGSTEN_BIN/thl -service parallel list' output
contains only statements which use the databases from
replicator.filter.replicate.do.

Is there some configuration/startup steps i'm missing, or this kind of
setup is not currently supported by tungsten-replicator?