hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jerry He (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-10367) RegionServer graceful stop / decommissioning
Date Fri, 20 Oct 2017 06:08:00 GMT

    [ https://issues.apache.org/jira/browse/HBASE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16212231#comment-16212231
] 

Jerry He commented on HBASE-10367:
----------------------------------

Yes, the AC hooks are added.

For the recommission, the region list is optional. (I tried to use the Optional, but in the
other JIRA we try to avoid using Optional as parameter.)  It is up to the user/caller.  If
it is not provided, then no region move.  decommission and recommission need to pair up.
But a normal graceful stop sequence would be:
0. get regions for the region server.
1. decommission call with offloading of the regions.
2. Stop the region server. 
3. Start the region server (after patching, e.g.).
4. recommission call with the last list of regions.

The current graceful-stop script does similar steps, but with manual region mover.  We can
get it to use the new APIs.

> RegionServer graceful stop / decommissioning
> --------------------------------------------
>
>                 Key: HBASE-10367
>                 URL: https://issues.apache.org/jira/browse/HBASE-10367
>             Project: HBase
>          Issue Type: Improvement
>            Reporter: Enis Soztutar
>            Assignee: Jerry He
>             Fix For: 3.0.0, 2.0.0-alpha-4
>
>         Attachments: HBASE-10367-master-2.patch, HBASE-10367-master.patch, HBASE-10367-master.patch
>
>
> Right now, we have a weird way of node decommissioning / graceful stop, which is a graceful_stop.sh
bash script, and a region_mover ruby script, and some draining server support which you have
to manually write to a znode (really!). Also draining servers is only partially supported
in LB operations (LB does take that into account for roundRobin assignment, but not for normal
balance) 
> See 
> http://hbase.apache.org/book/node.management.html and HBASE-3071
> I think we should support graceful stop as a first class citizen. Thinking about it,
it seems that the difference between regionserver stop and graceful stop is that regionserver
stop will close the regions, but the master will only assign them after the znode is deleted.

> In the new master design (or even before), if we allow RS to be able to close regions
on its own (without master initiating it), then graceful stop becomes regular stop. The RS
already closes the regions cleanly, and will reject new region assignments, so that we don't
need much of the balancer or draining server trickery. 
> This ties into the new master/AM redesign (HBASE-5487), but still deserves it's own jira.
Let's use this to brainstorm on the design. 



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Mime
View raw message