@Jarekt is this working for you now? Ok to close?

... and @Pchelolo and @Jdforrester-WMF !

+1 thanks @Eevans

Same as T244654 and T232087

AFAICS this is the same issue as T232087

Hah yeah, I've since realised that they don't work with our UI, nor should they. Raised T245264 which is related

FWIW 340 DivergingEntityIdExceptions in logstash in the last 7 days

Not happening on my local vagrant either :/

Thanks @Krinkle

@ArielGlenn the structured data team keep coming up as blockers on this ticket in scrum-of-scrums, but we're not blocking you, are we?

Great :)

@ArielGlenn ok to close this?

FYI this isn't just a config change, we don't have code that formats time or globe-coordinate values for the search index.

It's the one I always use :) I think we have some herald rules set up for some tags, but I'm not sure exactly which ones ... might ask you about it at all-hands

done

@Ramsey-WMF this isn't an Android team issue, it's a mobile web issue. @Niedzielski @Jdlrobson any ideas? We thought this might fix it but it didn't

@Jarekt merged a patch for this today https://gerrit.wikimedia.org/r/c/mediawiki/extensions/WikibaseMediaInfo/+/565308

Looks like we're getting ~500 File page creations a day, so we're talking at least 1000 delayed items being in the queue at all times. Is that a lot?

Thanks Max!

Aha! I found another flag --ignore-missing which I think does exactly what you need

AFAICS error msgs are written to php://stderr by default which I would expect to end up in logstash. There's a --log option in dumpRdf.php which allows you to specify where error msgs get written to, so I think you can probably send them to a file instead of logstash. Would that work for you? Probably you ought to test this first with --limit (I don't have logstash locally so I can't be sure I'm right)

Merged, so resolving

@brennen is this ok now?

There's a user preference to restore the original behaviour. Sorry that we neglected to mention it in the original announcement of this change

Ace, thanks @WMDE-leszek