On Fri, Feb 8, 2013 at 1:26 AM, Marcos Ortiz <[email protected]> wrote: > Regards, So, > Can you provide more information about your setup? > - HBase version > - Hadoop version > - Operating System > - Java version
This would be helpful. I would also like to see that GC log please. > Did you check the link? > Todd wrote a series of posts in Cloudera´s blog about Java Long GC pauses, > HBase and Zookeeper. > It´s a great read: > http://www.cloudera.com/blog/2011/02/avoiding-full-gcs-in-hbase-with-memstore-local-allocation-buffers-part-1/ > http://www.cloudera.com/blog/2011/02/avoiding-full-gcs-in-hbase-with-memstore-local-allocation-buffers-part-2/ The master doesn't have memstores so this wouldn't help. In fact it's pretty rare that we see the master with GC issues. I recall seing issues with time travelling (machine clock's too slow and ntpd resets it) or on EC2 where sometimes you'd see random machine pauses out of nowhere (although that was a long time ago and haven't used EC2 since). J-D
