[
https://issues.apache.org/jira/browse/PHOENIX-4785?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16514559#comment-16514559
]
Ankit Singhal commented on PHOENIX-4785:
----------------------------------------
bq. Yes, that's a shortcoming - I don't have any good idea to get around that.
[~vincentpoon], I think our comments were crossed, Please see my last comment.
bq. We are trying to strike a balance between not being overly aggressive in
disabling the index, and still have the index be as consistent as possible.
It's difficult to have both. I think the case of two clients both concurrently
writing AND one succeeds and the other fails AND one goes away , is a case we
have to compromise on for the time being.
The problem with this is it will make the index inconsistent silently. Consider
any distributed job(spark, storm, MR) writing data in Phoenix and admin made
some changes in HBase table by disabling and enabling it for a second, there is
always a possibility that we can run into this race condition.
If we can't fix it then either we should revert PHOENIX-4130 or make it
configurable (Disabled by default) so that user is allowed to make an informed
decision.
> Unable to write to table if index is made active during retry
> -------------------------------------------------------------
>
> Key: PHOENIX-4785
> URL: https://issues.apache.org/jira/browse/PHOENIX-4785
> Project: Phoenix
> Issue Type: Bug
> Affects Versions: 4.14.0
> Reporter: Romil Choksi
> Assignee: Vincent Poon
> Priority: Blocker
> Fix For: 5.0.0, 4.14.1
>
> Attachments: PHOENIX-4785.v1.master.patch, PHOENIX-4785_test.patch
>
>
> After PHOENIX-4130, we are unable to write to a table if an index is made
> ACTIVE during the retry as client timestamp is not cleared when table state
> is changed from PENDING_DISABLE to ACTIVE even if our policy is not to block
> writes on data table in case of write failure for index.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
