You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@phoenix.apache.org by "Nirmal (Jira)" <ji...@apache.org> on 2020/02/19 07:56:00 UTC
[jira] [Comment Edited] (PHOENIX-5731) Loading bulkload hfiles
should not be blocked if the upsert select happening for differet table.
[ https://issues.apache.org/jira/browse/PHOENIX-5731?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17039781#comment-17039781 ]
Nirmal edited comment on PHOENIX-5731 at 2/19/20 7:55 AM:
----------------------------------------------------------
The scenario explained in the issue could be described as while data is being ingested to _SOURCE_TABLE_ by "Application-1" using hfile bulk load approach using LoadIncrementalHFiles.doBulkLoad(...) if "Application-2" is started and that perform UPSERT into _DEST_TABLE_ SELECT from _SOURCE_TABLE_ then org.apache.hadoop.hbase.DoNotRetryIOException is noticed in "Application-1" impacting hfile bulk load operation.
was (Author: laddha):
The scenario explained in the issue could be described as while data is being ingested to _SOURCE_TABLE_ by "Application-1" using hfile bulk load approach using LoadIncrementalHFiles.doBulkLoad(...) if "Application-2" is started and that perform UPSERT into _DEST_TABLE_ SELECT from _SOURCE_TABLE_ then "Application-1" experiences org.apache.hadoop.hbase.DoNotRetryIOException.
> Loading bulkload hfiles should not be blocked if the upsert select happening for differet table.
> ------------------------------------------------------------------------------------------------
>
> Key: PHOENIX-5731
> URL: https://issues.apache.org/jira/browse/PHOENIX-5731
> Project: Phoenix
> Issue Type: Bug
> Reporter: Rajeshbabu Chintaguntla
> Assignee: Rajeshbabu Chintaguntla
> Priority: Major
> Fix For: 5.1.0, 4.16.0
>
> Attachments: PHOENIX-5731.patch
>
>
> currently we are not allowing to load hfiles after bulkload to avoid deadlock cases when upsert select is happening to same table but when the upserting to different table we should not block the load incremental hfiles.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)