Mailing-List: contact commits-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cassandra.apache.org
Date: Tue, 8 Aug 2017 06:26:02 +0000 (UTC)
From: "ASF GitHub Bot (JIRA)" <jira@apache.org>
To: commits@cassandra.apache.org
Message-ID: <JIRA.13085989.1499701312000.115105.1502173562557@Atlassian.JIRA>
In-Reply-To: <JIRA.13085989.1499701312000@Atlassian.JIRA>
References: <JIRA.13085989.1499701312000@Atlassian.JIRA> <JIRA.13085989.1499701312733@jira-lw-us.apache.org>
Subject: [jira] [Commented] (CASSANDRA-13683) sstabledump should not use
 tool initialization
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Tue, 08 Aug 2017 06:26:12 -0000


    [ https://issues.apache.org/jira/browse/CASSANDRA-13683?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16117921#comment-16117921 ] 

ASF GitHub Bot commented on CASSANDRA-13683:
--------------------------------------------

Github user asfgit closed the pull request at:

    https://github.com/apache/cassandra/pull/130


> sstabledump should not use tool initialization
> ----------------------------------------------
>
>                 Key: CASSANDRA-13683
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-13683
>             Project: Cassandra
>          Issue Type: Bug
>            Reporter: Chris Lohfink
>            Assignee: Chris Lohfink
>             Fix For: 4.0
>
>
> The tool initialization is not necessary for the sstabledump utility and by using it we possibly introduce a unnecessary requirement of being on the same C* instance (since we may introduce changes that require it, UDT changes can lead to dangerous assumptions) that has the schema. This also does things like introduce possibly devastating issues from having the same commitlog/sstable locations.
> A good example is updating the sstable activity table, and having that flushed to commitlogs as an application outside the running C* service. This may not be on same user (ie root) which than when a postmemtable flusher after restart attempts to delete it will throw an exception and potentially kill that thread pool. One that happens commitlogs will cease to get recycled and will just burn disk space until we run out.


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

---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cassandra.apache.org
For additional commands, e-mail: commits-help@cassandra.apache.org