[
https://issues.apache.org/jira/browse/IMPALA-14598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18113746#comment-18113746
]
Aleksandr Efimov edited comment on IMPALA-14598 at 9/10/26 8:53 AM:
--------------------------------------------------------------------
Hi [~stigahuang],
I have a practical use case for this in our Impala deployment.
I'm thinking about adding a simple local persistence mechanism that saves the
HBO cache to disk and restores it when the coordinator starts again. This
wouldn't share stats between coordinators, but it looks like a reasonable first
step.
Would it be useful for upstream?
was (Author: JIRAUSER313444):
Hi [~stigahuang],
I have a practical use case for this in our Impala deployment.
I'm thinking about adding a simple local persistence mechanism that saves the
HBO cache to disk and restores it when the coordinator starts again. This
wouldn't share stats between coordinators, but it looks like a useful first
step.
Would it be useful for upstream?
> Share historical stats between coordinators
> -------------------------------------------
>
> Key: IMPALA-14598
> URL: https://issues.apache.org/jira/browse/IMPALA-14598
> Project: IMPALA
> Issue Type: New Feature
> Reporter: Quanlong Huang
> Assignee: Quanlong Huang
> Priority: Critical
>
> Historical stats storing in memory can only be used by the current
> coordinator itself. Restarting the coordinator also loss the historical
> stats. We can store historical stats in a shared cache, e.g. Redis, to
> improve this.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
