Daniel Leffel wrote:
I've continued down the path of investigation here and after upgrading to
Hadoop 0.16.3 and Hbase 0.1.1, I've discovered the Region Server process is
dying with an OutOfMemory exception.
I've tried upping the HBase heap on the region server to 2 gigs.
This is happening during a TableReduce.
Are there any other setting I can tweak to help let the region server catch
up before running out of memory?
Give me a second. I'm trying to put together a 0.1.2 candidate. It has
backport of HBASE-512 which seems to make a big difference regards count
of regions any one server can carry. Once its up, take it for a spin.
St.Ack
