[ 
https://issues.apache.org/jira/browse/PHOENIX-7503?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Sanjeet Malhotra reassigned PHOENIX-7503:
-----------------------------------------

    Assignee:     (was: Sanjeet Malhotra)

> Enable compaction during flushes
> --------------------------------
>
>                 Key: PHOENIX-7503
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-7503
>             Project: Phoenix
>          Issue Type: Bug
>    Affects Versions: 5.2.1, 5.3.0
>            Reporter: Sanjeet Malhotra
>            Priority: Major
>
> PHOENIX-7402 removed compaction during flushes by removing override for 
> preFlush hook. The reason for removing the compaction from running during 
> flushes was retrieval of max lookback during preFlush hook execution. 
> Accessing SYSCAT from preFlush hook was problematic. Alternative is we can 
> store max lookback as a table attribute in HBase table and access that from 
> inside preFlush hook.
> In our production, we saw an instance where rows in a table are getting 
> updated very frequently and the rows were being updated partially so, due to 
> lack of compaction during flushes a lot of unwanted versions are being 
> retained during flushes and till first minor compaction.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to