You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues-all@impala.apache.org by "ASF subversion and git services (Jira)" <ji...@apache.org> on 2026/09/29 21:01:00 UTC

[jira] [Commented] (IMPALA-14686) Enable long polling by default

    [ https://issues.apache.org/jira/browse/IMPALA-14686?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18120899#comment-18120899 ] 

ASF subversion and git services commented on IMPALA-14686:
----------------------------------------------------------

Commit 446ea6afcab91d9a7bc1a19f986553c6e9b67342 in impala's branch refs/heads/master from Csaba Ringhofer
[ https://gitbox.apache.org/repos/asf?p=impala.git;h=446ea6afc ]

IMPALA-14177: Fix test_client_fetch_time_stats(_incomplete) on ARM

On systems where Impala uses CLOCK_MONOTONIC_COARSE the tests nearly
always failed since enabling long polling by default (IMPALA-14686):
the first "inactive period" took less than 1ms, which was counted as 0
with a ms precision clock, leading to ignoring the first request in
client_wait_time_stats_.

The coarse clock was used in the clocksource != "tsc" case, which
corresponds to ARM VMs in case of current Impala tests, but is not
specific to ARM in general.

Changes:
- always update client_wait_time_stats_ after the timer is started
  (it is possible to get a request before it starts, which is
  intentionally ignored in the counter: IMPALA-10180)
- added sleeps to tests to be more deterministic
- fix parsing logic in the tests (get_time_summary_stats_counter)
  which parsed Sum's value as Max due to greedy regexp
- close hs2 connection in tests when it fails - earlier the test
  failures also led to TestValidateMetrics.test_metrics_are_zero
  errors

Change-Id: I7a3b6d72109110c078abdae7483fb9f4c5df7f07
Assisted-by: Claude Opus 5.0
Reviewed-on: http://gerrit.cloudera.org:8080/24972
Reviewed-by: Arnab Karmakar <ar...@gmail.com>
Reviewed-by: Michael Smith <mi...@cloudera.com>
Reviewed-by: Joe McDonnell <jo...@cloudera.com>
Tested-by: Impala Public Jenkins <im...@cloudera.com>


> Enable long polling by default
> ------------------------------
>
>                 Key: IMPALA-14686
>                 URL: https://issues.apache.org/jira/browse/IMPALA-14686
>             Project: IMPALA
>          Issue Type: Improvement
>          Components: Clients, Frontend
>            Reporter: Michael Smith
>            Assignee: Joe McDonnell
>            Priority: Critical
>             Fix For: Impala 5.0.0
>
>
> IMPALA-13294 added support for long polling in Impala and impala-shell. This behavior is strictly better than the current behavior, allowing Impala clients to eliminate delays in completing show queries and making the experience of using Impala more responsive.
> On the fringes, clients that don't know about long polling may have slightly different timing with certain queries, as the RPC may wait the {{long_polling_time_ms}} timeout then do its own sleep.
> Set {{long_polling_time_ms}} to a short period - say 100ms - to provide better support for low-latency queries.



--
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