[jira] [Updated] (IMPALA-13684) waitForHmsEvent() should only wait for events on candidate tables

["Quanlong Huang (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/IMPALA-13684?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Quanlong Huang updated IMPALA-13684:
------------------------------------
    Issue Type: Bug  (was: Improvement)
       Summary: waitForHmsEvent() should only wait for events on candidate tables  (was: Improve waitForHmsEvent() to only wait for events on candidate tables)

> waitForHmsEvent() should only wait for events on candidate tables
> -----------------------------------------------------------------
>
>                 Key: IMPALA-13684
>                 URL: https://issues.apache.org/jira/browse/IMPALA-13684
>             Project: IMPALA
>          Issue Type: Bug
>          Components: Catalog
>            Reporter: Quanlong Huang
>            Assignee: Quanlong Huang
>            Priority: Critical
>             Fix For: Impala 5.0.0
>
>
> IMPALA-12152 adds a waiting mechanism before query planning to wait until catalogd is in synced with HMS events. It simply fetched the latest event id when the TWaitForHmsEventRequest is processed and wait until that event is processed.
> For non-transactional tables, we can leverage HIVE-27499 to efficiently detect whether the candidate tables have unsynced events and just wait for the *largest* id of them.
> On the other hand, if a candidate table is unloaded in catalogd, we don't need to wait for its events since there are no stale metadata of it cached in catalogd.
> CC [~hemanth619], [~VenuReddy]



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

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-all-unsubscribe@impala.apache.org
For additional commands, e-mail: issues-all-help@impala.apache.org