[
https://issues.apache.org/jira/browse/HBASE-4224?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13096799#comment-13096799
]
stack commented on HBASE-4224:
------------------------------
@Akash Thanks for volunteering. I don't think this one too hard to do. You'll
be able to figure it. Study what happens when you flush in the shell. See how
it passes a command to the master. In master, see how it fans it out by region
or by table. I think whats needed is a flush command that takes a machine
name. The shell code would figure distingush the flush of a region or table
from a flush of a regionserver (a bit of regexing should be able to distingush
whats wanted -- I'd imagine the flush a machine would take a ServerName; i.e.
<hostname> ',' <port> ',' <startcode>). The flush of a regionserver would
require adding a new API to HRegionInterface I'd imagine. It'd flush all
regions on the regionserver. Then on the end, after all flushed, it'd run the
check if we can let go of WAL logs. This method currently is checked on WAL
log roll. See in the HLog under the wal package. Ask questions if you help
figuring stuff. Thanks Akash.
> Need a flush by regionserver rather than by table option
> --------------------------------------------------------
>
> Key: HBASE-4224
> URL: https://issues.apache.org/jira/browse/HBASE-4224
> Project: HBase
> Issue Type: Bug
> Components: shell
> Reporter: stack
>
> This evening needed to clean out logs on the cluster. logs are by
> regionserver. to let go of logs, we need to have all edits emptied from
> memory. only flush is by table or region. We need to be able to flush the
> regionserver. Need to add this.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
