Mailing-List: contact issues-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Date: Thu, 3 Aug 2017 18:25:00 +0000 (UTC)
From: "Appy (JIRA)" <jira@apache.org>
To: issues@hbase.apache.org
Message-ID: <JIRA.13080560.1497670067000.87106.1501784700202@Atlassian.JIRA>
In-Reply-To: <JIRA.13080560.1497670067000@Atlassian.JIRA>
References: <JIRA.13080560.1497670067000@Atlassian.JIRA> <JIRA.13080560.1497670067987@jira-lw-us.apache.org>
Subject: [jira] [Commented] (HBASE-18231) Deprecate and throw unsupported
 operation when Admin#closeRegion is called.
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Thu, 03 Aug 2017 18:25:04 -0000


    [ https://issues.apache.org/jira/browse/HBASE-18231?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16113234#comment-16113234 ] 

Appy commented on HBASE-18231:
------------------------------

That'll break user clients.
From the "hbase 2.0 compatibility expectations" thread:
bq. An hbase1 client can run against an hbase2 cluster but it will only be able to do DML (Get/Put/Scan, etc.). We do not allow being able to do admin ops using an hbase1 Admin client against an hbase2 cluster. 

So we certainly don't have to worry about compat, but am not sure what's the accepted way to fail. Is it fine to just remove the methods, because then the clients might crash?
[~stack]

> Deprecate and throw unsupported operation when Admin#closeRegion is called.
> ---------------------------------------------------------------------------
>
>                 Key: HBASE-18231
>                 URL: https://issues.apache.org/jira/browse/HBASE-18231
>             Project: HBase
>          Issue Type: Sub-task
>          Components: Admin
>    Affects Versions: 2.0.0
>            Reporter: stack
>            Assignee: Appy
>            Priority: Critical
>             Fix For: 2.0.0, 3.0.0
>
>         Attachments: HBASE-18231.master.001.patch, HBASE-18231.master.002.patch, HBASE-18231.master.003.patch
>
>
> [~uagashe] tripped over this today. Admin#closeRegion which we used to use in branch-1 will cause damage in AMv2 cluster. Instead you need to call unassign -- i.e. all cluster ops must go via the Master; no more going direct to RegionServer closing regions behind the Master's back.
> Review all Admin ops to see what else skirts Master and deprecate and throw unsupported if called.


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