[
https://issues.apache.org/jira/browse/PHOENIX-4214?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16168806#comment-16168806
]
Vincent Poon commented on PHOENIX-4214:
---------------------------------------
[~lhofhansl] Yes, this patch blocks only new write requests, and lets existing
ones finish. I was thinking for case of local index creation, maybe we could
write to HFiles directly, then bulk load into HBase at the end? This would
avoid the flush/split deadlock issue in PHOENIX-3111 altogether. And HBase
1.3+ supports replication of bulk loads, I believe.
Normal upsert select and deletes would still need this patch, though.
> Scans which write should not block region split or close
> --------------------------------------------------------
>
> Key: PHOENIX-4214
> URL: https://issues.apache.org/jira/browse/PHOENIX-4214
> Project: Phoenix
> Issue Type: Bug
> Affects Versions: 4.12.0
> Reporter: Vincent Poon
> Attachments: splitDuringUpsertSelect_wip.patch
>
>
> PHOENIX-3111 introduced a scan reference counter which is checked during
> region preSplit and preClose. However, a steady stream of UPSERT SELECT or
> DELETE can keep the count above 0 indefinitely, preventing or greatly
> delaying a region split or close.
> We should try to avoid starvation of the split / close request, and
> fail/reject queries where appropriate.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
