Change 762054 had a related patch set uploaded (by Urbanecm; author: Urbanecm):

[mediawiki/extensions/GrowthExperiments@master] Mentor dashboard: Exclude globally locked users

https://gerrit.wikimedia.org/r/762054

Change 762054 merged by jenkins-bot:

[mediawiki/extensions/GrowthExperiments@master] Mentor dashboard: Exclude globally locked users

https://gerrit.wikimedia.org/r/762054

@Urbanecm_WMF - two questions:

(1) I checked sitewide (not global blocing) blocking on betalabs and on testwiki wmf.6 - there was no updates on MentorDashboard on blocked users, not during the duration of a block not after (I checked after 24 hours ).

testwiki wmf.6 shows the correct update - the delay is more than two hours though.

Steps:

• have a user as a mentee
• block that user sitewide. When testing, IP blocking was not involved, expiration was set to 1 and (for another test) to 2 hours.
• Special:MentorDashboard will not reflect the blocked status of a user - not during the blocked time, not after the expiration of the block.

(2) Global blocking does have an option to Unblock, so a globally blocked user might be unblocked and, probably, it makes sense to keep it on the MentorDashboard?

Hello @Etonkovidova,

Thanks for your feedback. I'm not sure I understand though – perhaps I'm missing something important here. In this task, mentee overview module logic was changed to exclude users who are globally locked (concept of locks is different from both global blocks [applicable to IP addresses/ranges only] and local blocks [applicable to both accounts/IPs, but only at the wiki level]). The logic for number of blocks should be left unchanged.

(1) I checked sitewide (not global blocing) blocking on betalabs and on testwiki wmf.6 - there was no updates on MentorDashboard on blocked users, not during the duration of a block not after (I checked after 24 hours ).

Does this mean that the number of blocks in the last column did not increment? That'd be weird. I tested it at https://ctrlv.tv/bLbe, and it looks to work, Granted, I did not wait three hours and I invoked the update manually from the maintenance server instead, to save some time, so it's still possible the scheduling is broken for some reason. I can look at that if that's the issue you noticed.

testwiki wmf.6 shows the correct update - the delay is more than two hours though.

Was it between two and three hours? If so, it probably works correctly, as the update is supposed to happen every three hours. Precisely speaking, update is started at the following times:

• 00:15:00 UTC
• 03:15:00 UTC
• 06:15:00 UTC
• 09:15:00 UTC
• 12:15:00 UTC
• 15:15:00 UTC
• 18:15:00 UTC
• 21:15:00 UTC

The update can take a couple of minutes to finish though. /var/log/mediawiki/mediawiki_job_growthexperiments-updateMenteeData-* at the maintenance server (mwmaint1002.eqiad.wmnet for production, deployment-mwmaint02.deployment-prep.eqiad1.wikimedia.cloud for beta) has logs of the process, and it can be used to check whether it ran already.

(2) Global blocking does have an option to Unblock, so a globally blocked user might be unblocked and, probably, it makes sense to keep it on the MentorDashboard?

Note that global blocks can be only applied to IP addresses, not users. If an account is unlocked (their lock is removed), they should re-appear in the dashboard (so long they meet the other conditions, like age of their last edit). This should also work for local blocks.

Let me know if I understood anything you said incorrectly.