You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@hbase.apache.org by ns...@apache.org on 2011/10/11 04:03:49 UTC

svn commit: r1181381 - /hbase/branches/0.89/bin/hbase-config.sh

Author: nspiegelberg
Date: Tue Oct 11 02:03:49 2011
New Revision: 1181381

URL: http://svn.apache.org/viewvc?rev=1181381&view=rev
Log:
Increase RS Timeout

Summary:
HBASE_SLEEP_TIMEOUT is used to determine that an RS connection is hosed so you
can move to the next one.  For our uber-large, uber-loaded clusters, a legit
stop can take 1-2 minutes.  Raise this threshold.  Long-term solution = bound
close time in RS code.

Test Plan:
bin/rolling-restart.sh

DiffCamp Revision: 160116
Reviewed By: kannan
CC: kannan
Revert Plan:
OK

Modified:
    hbase/branches/0.89/bin/hbase-config.sh

Modified: hbase/branches/0.89/bin/hbase-config.sh
URL: http://svn.apache.org/viewvc/hbase/branches/0.89/bin/hbase-config.sh?rev=1181381&r1=1181380&r2=1181381&view=diff
==============================================================================
--- hbase/branches/0.89/bin/hbase-config.sh (original)
+++ hbase/branches/0.89/bin/hbase-config.sh Tue Oct 11 02:03:49 2011
@@ -148,3 +148,5 @@ fi
 if [ -z "$HBASE_ZOOKEEPER_OPTS" ]; then
   export HBASE_ZOOKEEPER_OPTS="$HBASE_ZOOKEEPER_DBG_OPTS $HBASE_ZOOKEEPER_JMX_OPTS"
 fi
+
+export HBASE_SLAVE_TIMEOUT=300