[
https://issues.apache.org/jira/browse/HBASE-10896?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13985187#comment-13985187
]
Mikhail Antonov commented on HBASE-10896:
-----------------------------------------
My point is - I think process name should be not just informative, but also
parseable by scripts (like _jps | grep blah-blah_)
Something like:
- hbase_rs - simple RS
- hbase_m - simple active master
- hbase_mrs - active master + rs
You think backup master we don't need to display specifically? (also at some
point.. we would have multiple active masters I guess, so backups would
naturally go away :))
I can take this one.
> Listing processes, master+regionserver shows as HMaster, others as
> HRegionServer
> --------------------------------------------------------------------------------
>
> Key: HBASE-10896
> URL: https://issues.apache.org/jira/browse/HBASE-10896
> Project: HBase
> Issue Type: Task
> Reporter: stack
>
> If I do a process listing I see the combined Master and RegionServer as
> follows:
> 30386 HMaster
> ... and the regionserver only process as:
> 16895 HRegionServer
> Backup masters though they are serving regions show as 'HMaster'.
> Should process now be called 'HBase' instead?
> IntegratonTests IIRC, might kill based off process name so changing process
> name would mess this up.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
