One aspect to note w.r.t. using PrefixFilter is that the start row is not automatically set. You need to pass the correct start row along with PrefixFilter. Otherwise the number of rows scanned may be quite high.
... View more
ACLs were not part of the core hbase (implemented through coprocessor). We were adding to a core functionality knowledge about an external component (ACL). There was the discussion about the meaning of restoring the acls that we snapshotted (see HBASE-11013). Please consult with the following for up-to-date ACL: http://hbase.apache.org/book.html#appendix_acl_matrix We cannot allow any user to restore any snapshot otherwise you'll be able to see data that is not yours. There is no ACL on snapshot to say "allow this user to restore/clone" the snapshot.
... View more
You should spread out the major compaction on individual tables. Refer to the following: HBASE-16147 Add ruby wrapper for getting compaction state If there is not enough time to wait for compaction on one table to finish, at least leave some time between the start of major compactions.
... View more
Did you mean locality for HFiles ? You can get such metric through /jmx There isn't command line which would directly give you this metric.
... View more
Please take a look at: https://issues.apache.org/jira/browse/HBASE-11339 which would reduce I/O amplification incurred by medium objects. This feature is in the upcoming HDP 2.5 release.
... View more
By default, each component (NodeManager, Datanode, and RegionServer daemon) would have their own predetermined heap memory setting. You need to tune yourself.
... View more
Can you find encoded region name for regions stuck in FAILED_OPEN state and pastebin related region server log ? It would help us understand what caused the region not to open. BTW which HDP version are you running ?
... View more
2016-07-06 14:59:14,466 INFO org.apache.zookeeper.ZooKeeper: Initiating client connection, connectString=m2.domain:2181 sessionTimeout=120000 watcher=hconnection- 0x7bc9e6ab0x0, quorum=m2.domain:2181, baseZNode=/hbase-secure Looks like AMS tried to connect to hbase cluster's znode. AMS should use /ams-hbase-secure as base znode. Can you check your configuration ?
... View more
