Hi Ninad, Thanks for your comments.
We have already committed 3G memory to each of the region servers. And the following texts are excerpted from hbase-env.sh in our cluster. # The java implementation to use. Java 1.6 required. export JAVA_HOME=/usr/lib/jvm/java-6-sun # The maximum amount of heap to use, in MB. Default is 1000. export HBASE_HEAPSIZE=3000 # Extra Java runtime options. Empty by default. export HBASE_OPTS="-XX:+UseConcMarkSweepGC -XX:+CMSIncrementalMode -verbose:gc -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -Xloggc:/opt/hbase/logs/gc-hbase.log" On Mon, Jul 27, 2009 at 6:03 PM, Ninad Raut<[email protected]> wrote: > I think there is a swap happening on regionservers due to less memory... > increase the allotted memory in hadoop-site.xml for all the slaves. > > On Mon, Jul 27, 2009 at 3:19 PM, Angus He <[email protected]> wrote: > -- Regards Angus
