Hello there. I have set-up this event[1] to pull data for 28 usernames. It error's out when I try to pull data for more than 1 day. What could be wrong?
Link:
[1] https://eventmetrics.wmflabs.org/programs/618/events/2098
Description
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Stalled | None | T285788 Event to pull data for 28 usernames errors when trying to pull data for more than 1 day | |||
| Open | BUG REPORT | None | T286328 Simple query scans entire revision table on new replicas |
Event Timeline
Comment Actions
Normally when this happens it's because you're querying against a very large number of revisions, for instance if you included a very prolific user as a participant (who has made many millions of edits), or if you use a very long date range. This does not seem to be the case here. Your event should process just fine.
Investigating further, I am now convinced there's an issue with the replica databases. The architecture of the databases were recently redesigned so it's possible that's related. I have filed a bug at T286328.
Unfortunately, I don't have a workaround for you right now. It's possible if you remove some participants or shorten the time frame it will finish, but I can't say for sure. The conditions which cause a timeout are seemingly arbitrary. In fact, adding more participants or broadening the date range might even help. By all indications, something is wrong on the database side of things. Thanks for letting us know!