1722287 - 2.75 - 0.85% Heap Unclassified / Explicit Memory + 1 more (Linux, Windows) regression on Mon July 19 2021

Status: RESOLVED WONTFIX

(Core :: Networking: HTTP, defect)

Description

[Acasandrei Beatrice (needinfo me)]

Perfherder has detected a awsy performance regression from push b633563955ad437d61cb5ddb11b39074764c8e6d. As author of one of the patches included in that push, we need your help to address this regression.

Regressions:

Ratio	Suite	Test	Platform	Options	Absolute values (old vs new)
3%	Heap Unclassified		windows10-64-shippable-qr		48,726,904.02 -> 50,068,264.42
3%	Heap Unclassified		windows10-64-shippable-qr		48,811,843.84 -> 50,073,676.53
1%	Explicit Memory		linux1804-64-shippable		353,922,209.40 -> 356,948,007.61

Details of the alert can be found in the alert summary, including links to graphs and comparisons for each of the affected tests. Please follow our guide to handling regression bugs and let us know your plans within 3 business days, or the offending patch(es) will be backed out in accordance with our regression policy.

For more information on performance sheriffing please see our FAQ.

Comment 1

[Acasandrei Beatrice (needinfo me)]

We are also seeing this alert on mozilla-beta:

== Change summary for alert #30633 (as of Thu, 22 Jul 2021 18:19:54 GMT) ==

Regressions:

Ratio	Suite	Test	Platform	Options	Absolute values (old vs new)
5%	Heap Unclassified		windows10-64-shippable-qr		48,573,022.68 -> 50,961,877.09
3%	Heap Unclassified		linux1804-64-shippable		56,749,494.28 -> 58,369,095.20
2%	Resident Memory		windows10-64-shippable-qr		537,771,620.04 -> 550,014,888.92
2%	Explicit Memory		windows10-64-shippable-qr		361,606,762.42 -> 369,343,812.13
2%	Explicit Memory		linux1804-64-shippable		361,402,733.62 -> 368,840,898.70

For up to date results, see: https://treeherder.mozilla.org/perfherder/alerts?id=30633

Comment 3

[Shane Caraveo (:mixedpuppy)]

I'm looking into this.

Comment 4

[Shane Caraveo (:mixedpuppy)]

This is still being actively investigated, am pulling more people into looking at this. It really doesn't appear that these changes could causes this kind of change.

Comment 5

[Pascal Chevrel:pascalc (PTO until April 26)]

We can target a fix for 92, Wontfix for 91.

Comment 6

[Shane Caraveo (:mixedpuppy)]

We're going to live with this for 91 and make an improvement in 92

Comment 7

[Shane Caraveo (:mixedpuppy)]

Additional patches have landed on autoland today that should minimize the potential number of requests affected by the earlier change. Hopefully we'll be able to see some change reflected in the coming days.

Comment 8

[Ryan VanderMeulen [:RyanVM]]

I'm having a hard time finding graphs of recent data to indicate whether the changes in bug 1722942 helped improve this or not. Can you please help, Beatrice?

Comment 9

[Acasandrei Beatrice (needinfo me)]

(In reply to Ryan VanderMeulen [:RyanVM] from comment #8)

I'm having a hard time finding graphs of recent data to indicate whether the changes in bug 1722942 helped improve this or not. Can you please help, Beatrice?

It seems that the tests mentioned above in the regressions summary stopped running in August. We'll investigate if that is expected or not.

Comment 10

[Valentin Gosu [:valentin] (he/him)]

(In reply to Acasandrei Beatrice (needinfo me) from comment #9)

(In reply to Ryan VanderMeulen [:RyanVM] from comment #8)

I'm having a hard time finding graphs of recent data to indicate whether the changes in bug 1722942 helped improve this or not. Can you please help, Beatrice?

It seems that the tests mentioned above in the regressions summary stopped running in August. We'll investigate if that is expected or not.

Was this resolved in any way. Is this bug still actionable?

Comment 11

[Acasandrei Beatrice (needinfo me)]

Hi. I don't think it's still actionable, It's status can be changed to WONTFIX.

Comment 12

[Valentin Gosu [:valentin] (he/him)]

Thank you!