[jira] [Commented] (IMPALA-7086) Cache timezone name look ups

["Mihaly Szjatinya (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/IMPALA-7086?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17877692#comment-17877692 ] 

Mihaly Szjatinya commented on IMPALA-7086:
------------------------------------------

[https://github.com/apache/impala/pull/79]

 

Added Prepare - Close routine arount from/to_utc standard functions. This gives pretty consistent time improvement for constant timezones.

 

Given sample table with 600M timestamp rows, on all-default environment the query below gives a stable 2-3 seconds improvement.
{code:java}
SELECT count(*) FROM timestamp_test2 where from_utc_timestamp(ts, "Europe/Budapest") > "2025-01-01";{code}

> Cache timezone name look ups
> ----------------------------
>
>                 Key: IMPALA-7086
>                 URL: https://issues.apache.org/jira/browse/IMPALA-7086
>             Project: IMPALA
>          Issue Type: Improvement
>          Components: Backend
>            Reporter: Csaba Ringhofer
>            Assignee: Mihaly Szjatinya
>            Priority: Major
>              Labels: performance, ramp-up, timestamp
>
> to/from_utc_timestamp looks up time zones by name during every invocation, even if the timezone parameter is constant. Avoiding this lookup if time zone name is the same as during the last call (in the fragment) could speed up time zone conversions.



--
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