Page MenuHomePhabricator

VirtualPageViews instrumentation broken in 1.37.0-wmf.17
Closed, ResolvedPublicBUG REPORT

Description

VirtualPageViews is part of a core metric but stopped logging with last train's deploy due to a rebase error in T267211.
https://grafana.wikimedia.org/d/000000566/overview?viewPanel=6&orgId=1

This should be fixed by https://gerrit.wikimedia.org/r/c/mediawiki/extensions/Popups/+/710724 and should be immediately backported.

Note: Alerts have been setup to avoid this happening in future.

Event Timeline

Jdlrobson triaged this task as Unbreak Now! priority.Aug 11 2021, 8:00 PM

Change 710724 had a related patch set uploaded (by Jdlrobson; author: Phuedx):

[mediawiki/extensions/Popups@wmf/1.37.0-wmf.18] virtualPageView: Log VirtualPageView events to Event Platform

https://gerrit.wikimedia.org/r/710724

Change 710723 had a related patch set uploaded (by Jdlrobson; author: Phuedx):

[mediawiki/extensions/Popups@wmf/1.37.0-wmf.17] virtualPageView: Log VirtualPageView events to Event Platform

https://gerrit.wikimedia.org/r/710723

Change 710724 merged by jenkins-bot:

[mediawiki/extensions/Popups@wmf/1.37.0-wmf.18] virtualPageView: Log VirtualPageView events to Event Platform

https://gerrit.wikimedia.org/r/710724

Change 710723 merged by jenkins-bot:

[mediawiki/extensions/Popups@wmf/1.37.0-wmf.17] virtualPageView: Log VirtualPageView events to Event Platform

https://gerrit.wikimedia.org/r/710723

Mentioned in SAL (#wikimedia-operations) [2021-08-11T20:20:20Z] <mholloway-shell@deploy1002> Synchronized php-1.37.0-wmf.18/extensions/Popups: Log VirtualPageView events to Event Platform (T288655) (duration: 01m 09s)

Mentioned in SAL (#wikimedia-operations) [2021-08-11T20:23:26Z] <mholloway-shell@deploy1002> Synchronized php-1.37.0-wmf.17/extensions/Popups: Log VirtualPageView events to Event Platform (T288655) (duration: 01m 06s)

Jdlrobson claimed this task.
Jdlrobson added subscribers: jlinehan, phuedx, Milimetric.

Big thanks to @Mholloway @jlinehan @phuedx and @Milimetric for identifying the issue

The graph has recovered after the backport:
.

Screen Shot 2021-08-11 at 1.35.29 PM.png (918×2 px, 190 KB)

I'll check in ticket T267211 after a few days to make sure we are seeing normal levels.

I've documented this in https://www.mediawiki.org/wiki/Reading/Web/Notable_incidents and set up an email alert to avoid us going so long without noticing such an issue again.