Confirmed and triaging to normal because all pages should show up.

This page shows up now. This problem could have been caused by the general brokenness we had in CirrusSearch over the past few days that is now resolved. It could, on the other hand, be real and masked by the first that we used to "catch up" on things we missed after the problem.

I'm keeping this open so I can add a regression test to CirrusSearch for these very pages. If it passes I'll call this a problem with the search brokenness and be happy. Otherwise it is a problem in the code and this'll catch it and I'll fix it.

Also, another thing to keep in mind is that the cache time for those searches is very very long, still. That caching is both done at WMF and in your browser. I'm not sure if that is a problem or not because searching for new pages will bring them up because the results won't be cached.

Adding "all" subpages to the subject. In my own tests it may happen that one or more subpages are listed, but not all.

Esther (and I) can see the ESM/proves subpages, which is apparently the last one she created. Maybe this has something to do with new subpages being indexed while old subpages not?

Ah, indeed, now I can see several subpages while yesterday therewas only one: /proves.

Esther, can you test again and report whether you see now all subpages in the search suggestions?

Quim,

Can you run through the tests again and post the failing pages to this bug? The search brokenness really could have invalidated results from before about an hour ago, unfortunately. cawiki is big and took a long time to recover.

Not that I'm happy with this problem, I just want to figure out if this is an ongoing issue or a symptom of one we've already solved.

Thanks for checking this issue. Apparently, my user subpages are displayed if I run the search with Safari (I haven't been using it for ages). OTOH, after cleaning the cache in Firefox, the issue persists and I only see Usuari:ESM/proves

Don't know what can be wrong. Thanks for your help!

(In reply to comment #5)

Quim,

Can you run through the tests again and post the failing pages to this bug?

I tried with

Usuari:ESM/ (the reporter)
Usuari:QuimGil/ (your servant, because I know the subpages I have)
Usuari:KRLS/ (an admin with a lot of subpages)

All seem to work fine now in my Chrome and Firefox browsers.

Esther, as The IT Crowd says: "Have you turned it off and on"? :) (It's funny because I'm serious).

Dang, it seems to be perfectly tuned now. Thanks for your time!

Sorry for it being broken at the time! I'll still keep this open to add the regression test.