I tested it again very carefully and found some interesting aspects : > 3 brokers were running fine where one is master and other two are slaves. > I stopped one of the slaves and deleted all leveldb files. > I copied all master broker leveldb files into the slaves leveldb > directory. (removed nodeid.txt) > Now when i start primary, I am seeing that > One tmp folder is created which has some .log files. > Now slave will download data log files from master and as soon as it copies all the log files, it is deleting the stash folder and i guess replaying in indexes from logs. > After that everything works well./
Snapshot of stash directory: 14:59]khand...@agni05.hyd:/u/khandela/activemq/amq3>du -sh leveldb/* 101M leveldb/0000000000000000.log 952K leveldb/00000000064004c6.log 4.0K leveldb/dirty.index 0 leveldb/lock 0 leveldb/nodeid.txt *13M leveldb/stash* 0 leveldb/store-version.txt and [14:59]khand...@agni05.hyd:/u/khandela/activemq/amq3>du -sh leveldb/stash/* 101M leveldb/stash/0000000000000000.log 0 leveldb/stash/0000000000000000.log.crc32 5.1M leveldb/stash/00000000064004c6.index 7.2M leveldb/stash/00000000064004c6.log Activemq vendors, please respond about above observations. Thanks, Anuj -- View this message in context: http://activemq.2283324.n4.nabble.com/Replicated-LevelDB-Manually-copying-data-from-one-leveldb-store-to-another-tp4686931p4686950.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.
