[jira] [Resolved] (IMPALA-7118) Add mem_limit metric listing for all hosts in the profile

["Tim Armstrong (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/IMPALA-7118?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Tim Armstrong resolved IMPALA-7118.
-----------------------------------
    Resolution: Won't Fix

I think this is of limited use if you don't also know how much memory other queries are consuming on the backends. Not convinced that the profile is the right place for this.

IMPALA-7811 added this to the /backends debug page and IMPALA-8162 added the current amount of memory consumed, which should make it easier to see problems without having to inspect all the nodes.

> Add mem_limit metric listing for all hosts in the profile
> ---------------------------------------------------------
>
>                 Key: IMPALA-7118
>                 URL: https://issues.apache.org/jira/browse/IMPALA-7118
>             Project: IMPALA
>          Issue Type: Improvement
>          Components: Backend
>    Affects Versions: Impala 2.10.0, Impala 2.12.0
>            Reporter: Mala Chikka Kempanna
>            Priority: Trivial
>              Labels: observability, ramp-up, supportability
>
> Similar to peak node memory metric in profile like below, we also need a metric listing mem_limit set on each node in profile.
> {code:java}
> Per Node Peak Memory Usage: x.y1.com:22000(36.14 GB) 
> {code}
>  
> Background for this ask is:
> We have seen performance issues reported post-upgrade of impala, where cluster admin would have lowered memory limits on all or subset of impala daemons while also upgrading the version.
> And query performance suffers if the queries that were running in-memory before start spilling now due to lower memory limit.
> And end-users/application-users blame the upgraded version for slowness when they share profiles for analysis, not knowing the lowering of memory limits for impala by cluster-admins was the reason.
> Having this mem limit metric in profile will help easily spot this issue.
>  
>  
>  



--
This message was sent by Atlassian Jira
(v8.3.4#803005)