[
https://issues.apache.org/jira/browse/HBASE-10688?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Enis Soztutar updated HBASE-10688:
----------------------------------
Attachment: 0001-HBASE-10688-v1.patch
Attaching a patch for a ruby script for managing draining_nodes in zk:
{code}
Usage: ./hbase org.jruby.Main draining_servers.rb [options] add|remove|list
<hostname>|<host:port>|<servername> ...
Add remove or list servers in draining mode. Can accept either hostname to
drain all region serversin that host, a host:port pair or a host,port,startCode
triplet. More than one server can be given separated by space
-h, --help Display usage information
-d, --debug Display extra debug logging
{code}
I have tested this with:
{code}
bin/hbase org.jruby.Main bin/draining_servers.rb add localhost,60010,23223
10.11.2.75,49740
bin/hbase org.jruby.Main bin/draining_servers.rb list
localhost,60010,23223
10.11.2.75,49740,1391051351778
bin/hbase org.jruby.Main bin/draining_servers.rb remove 10.11.2.75,49740
bin/hbase org.jruby.Main bin/draining_servers.rb list
localhost,60010,23223
{code}
> Add a draining_node script to manage nodes in draining mode
> -----------------------------------------------------------
>
> Key: HBASE-10688
> URL: https://issues.apache.org/jira/browse/HBASE-10688
> Project: HBase
> Issue Type: Sub-task
> Components: scripts
> Reporter: Enis Soztutar
> Assignee: Enis Soztutar
> Fix For: 0.99.0
>
> Attachments: 0001-HBASE-10688-v1.patch
>
>
> As per parent issue, our support for graceful stop and node decommissioning
> can be improved a lot. However, until we fix it for good, we can add a
> draining_node script to manage the nodes in draining mode in zk.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
