[
https://issues.apache.org/jira/browse/HIVE-20943?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Eugene Koifman updated HIVE-20943:
----------------------------------
Attachment: HIVE-20943.04.patch
> Handle Compactor transaction abort properly
> -------------------------------------------
>
> Key: HIVE-20943
> URL: https://issues.apache.org/jira/browse/HIVE-20943
> Project: Hive
> Issue Type: Bug
> Components: Transactions
> Affects Versions: 4.0.0
> Reporter: Eugene Koifman
> Assignee: Eugene Koifman
> Priority: Major
> Attachments: HIVE-20943.01.patch, HIVE-20943.02.patch,
> HIVE-20943.02.patch, HIVE-20943.03.patch, HIVE-20943.03.patch,
> HIVE-20943.04.patch
>
>
> A transactions in which the Worker runs may fail after base_x_cZ
> (delta_x_y_cZ) is created but before files are fully written. Need to make
> sure to write to TXN_COMPONENTS an entry for corresponding to Z so "_cZ"
> directories are not read by anyone and cleaned by Cleaner.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
