You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@phoenix.apache.org by "Thomas D'Silva (JIRA)" <ji...@apache.org> on 2018/10/11 04:44:00 UTC

[jira] [Comment Edited] (PHOENIX-4963) Do not throw when transitioning to PENDING_DISABLE if index is already disabled

    [ https://issues.apache.org/jira/browse/PHOENIX-4963?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16645969#comment-16645969 ] 

Thomas D'Silva edited comment on PHOENIX-4963 at 10/11/18 4:43 AM:
-------------------------------------------------------------------

+1

So if a client tries to change the index state from DISABLE to PENDING_DISABLE and we return the MutationCode.TABLE_ALREADY_EXISTS that means an exception won't be thrown? 



was (Author: tdsilva):
So if a client tries to change the index state from DISABLE to PENDING_DISABLE and we return the MutationCode.TABLE_ALREADY_EXISTS that means an exception won't be thrown? 

> Do not throw when transitioning to PENDING_DISABLE if index is already disabled
> -------------------------------------------------------------------------------
>
>                 Key: PHOENIX-4963
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-4963
>             Project: Phoenix
>          Issue Type: Bug
>    Affects Versions: 4.14.0
>            Reporter: Vincent Poon
>            Assignee: Vincent Poon
>            Priority: Critical
>         Attachments: PHOENIX-4963.v1.master.patch
>
>
> When multiple clients get index write failures, they may attempt to transition to PENDING_DISABLE concurrently.  A race condition can cause one to exhaust its retries first and transition the index to DISABLE, and another client to then attempt to transition from DISABLE to PENDING_DISABLE.  Currently this is causing an exception to get thrown, leading to the KillServerOnFailurePolicy getting activated.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)