Hi, I'm using a hbase0.94.18 cluster with three region servers on Amazon AWS EMR. The table is spitted into several hundreds regions. And the queries are all random get from java client api.
I noticed that the number of 'waited count' of IPC server handler is always increasing, which also causes the request per second of my web service is really low(~2500). Is there any suggestion of approaching this problem? Thank you! -- *Thank you && Best Regards,* *Zhiyuan Yang*
