[ https://issues.apache.org/jira/browse/HBASE-2312?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12974029#action_12974029 ]
HBase Review Board commented on HBASE-2312: ------------------------------------------- Message from: "Nicolas" <nspiegelb...@facebook.com> ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://review.cloudera.org/r/396/ ----------------------------------------------------------- (Updated 2010-12-21 19:06:32.166768) Review request for hbase. Changes ------- Version for 0.90. This version utilizes a new HDFS patch to forcibly recover a file lease (forthcoming). TestZooKeeper will fail without this patch because it needs to wait until the soft lease expires otherwise. Summary ------- There is a very corner case when bad things could happen(ie data loss): 1) RS #1 is going to roll its HLog - not yet created the new one, old one will get no more writes 2) RS #1 enters GC Pause of Death 3) Master lists HLog files of RS#1 that is has to split as RS#1 is dead, starts splitting 4) RS #1 wakes up, created the new HLog (previous one was rolled) and appends an edit - which is lost Note that this fix requires a healthy dose of HDFS prerequisites: HDFS-617, HADOOP-6840, HADOOP-6886. I encourage you to review those as well, give feedback, and hopefully give +1s so we can push the changes through. This addresses bug HBASE-2312. http://issues.apache.org/jira/browse/HBASE-2312 Diffs (updated) ----- trunk/src/main/java/org/apache/hadoop/hbase/HConstants.java 1051398 trunk/src/main/java/org/apache/hadoop/hbase/master/MasterFileSystem.java 1051398 trunk/src/main/java/org/apache/hadoop/hbase/regionserver/wal/HLog.java 1051398 trunk/src/main/java/org/apache/hadoop/hbase/regionserver/wal/SequenceFileLogWriter.java 1051398 trunk/src/main/java/org/apache/hadoop/hbase/util/FSUtils.java 1051398 trunk/src/test/java/org/apache/hadoop/hbase/regionserver/wal/TestHLogSplit.java 1051398 Diff: http://review.cloudera.org/r/396/diff Testing ------- mvn test; bin/start-hbase.sh bin/hbase shell < scan '.META.', get, put, etc Thanks, Nicolas > Possible data loss when RS goes into GC pause while rolling HLog > ---------------------------------------------------------------- > > Key: HBASE-2312 > URL: https://issues.apache.org/jira/browse/HBASE-2312 > Project: HBase > Issue Type: Bug > Components: master, regionserver > Affects Versions: 0.90.0 > Reporter: Karthik Ranganathan > Assignee: Nicolas Spiegelberg > Priority: Critical > Fix For: 0.90.1 > > > There is a very corner case when bad things could happen(ie data loss): > 1) RS #1 is going to roll its HLog - not yet created the new one, old one > will get no more writes > 2) RS #1 enters GC Pause of Death > 3) Master lists HLog files of RS#1 that is has to split as RS#1 is dead, > starts splitting > 4) RS #1 wakes up, created the new HLog (previous one was rolled) and > appends an edit - which is lost > The following seems like a possible solution: > 1) Master detects RS#1 is dead > 2) The master renames the /hbase/.logs/<regionserver name> directory to > something else (say /hbase/.logs/<regionserver name>-dead) > 3) Add mkdir support (as opposed to mkdirs) to HDFS - so that a file > create fails if the directory doesn't exist. Dhruba tells me this is very > doable. > 4) RS#1 comes back up and is not able create the new hlog. It restarts > itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.