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

Posted to user@hbase.apache.org by Yves Langisch <yv...@langisch.ch> on 2012/01/21 21:29:27 UTC

Interpret metrics

Hi,

I have some performance issues and I try to interpret the HBase metrics I collect trough ganglia. I've noticed that the metric 'dfs.datanode.writeBlockOp_avg_time' is pretty high sometimes (~3'500'000) before dropping near to zero again. How do I have to interpret that characteristic?

Another question I have is about the 'hbase.regionserver.memstoreSizeMB' metric. My memstore size is growing constantly over a period of an hour up to 1GB. After that it drops down to 900MB and grows again up to 1GB and so on...I thought the memstore is completely flushed once it reaches its maximum size (just like here http://outerthought.org/blog/465-ot.html)? Why do I see a different behavior?

Thanks
Yves