You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "ASF GitHub Bot (Jira)" <ji...@apache.org> on 2022/05/20 06:49:00 UTC

[jira] [Work logged] (HIVE-26242) Compaction heartbeater improvements

     [ https://issues.apache.org/jira/browse/HIVE-26242?focusedWorklogId=772684&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-772684 ]

ASF GitHub Bot logged work on HIVE-26242:
-----------------------------------------

                Author: ASF GitHub Bot
            Created on: 20/May/22 06:48
            Start Date: 20/May/22 06:48
    Worklog Time Spent: 10m 
      Work Description: veghlaci05 opened a new pull request, #3303:
URL: https://github.com/apache/hive/pull/3303

   ### What changes were proposed in this pull request?
   This PR introduces a new component (CompactionHeartbeatService) to centralize the compaction transaction heartbeating, and reduce resource usage. It consist of a Scheduled exevutor service and a MetaStoreClient pool.
   
   ### Why are the changes needed?
   The heartbeat of the compaction txns are wasting resources by having a dedicated separate thread and MetaStore client for every compaction transaction.
   
   ### Does this PR introduce _any_ user-facing change?
   No
   
   
   ### How was this patch tested?
   Manually and through unit tests.




Issue Time Tracking
-------------------

            Worklog Id:     (was: 772684)
    Remaining Estimate: 0h
            Time Spent: 10m

> Compaction heartbeater improvements
> -----------------------------------
>
>                 Key: HIVE-26242
>                 URL: https://issues.apache.org/jira/browse/HIVE-26242
>             Project: Hive
>          Issue Type: Improvement
>            Reporter: László Végh
>            Priority: Major
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> The Compaction heartbeater should be improved the following ways:
>  * The metastore clients should be reused between heartbeats and closed only at the end, when the transaction ends
>  * Instead of having a dedicated heartbeater thread for each Compaction transaction, there should be shared a heartbeater executor where the heartbeat tasks can be scheduled/submitted.



--
This message was sent by Atlassian Jira
(v8.20.7#820007)