Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 158 Stack Exchange communities.

What is meta?
Here's how it works:
  1. Any Stack Exchange user can ask a question
  2. The community provides support, votes on ideas, and reports bugs
  3. Your voice helps shape the way Stack Exchange operates

Today my trusty search for newest zero-answer C# questions:

which is basically "answers:0 [c#]" and switch to the "newest" tab, returns questions starting nine hours ago and it includes questions with answers:

enter image description here

share|improve this question
[status-norepro] for me. Are you ignoring the C# tag? – Robert Harvey May 12 '11 at 23:08
No, not ignored, but it is a favorite. That causes the flesh tone background in the hits above. If I do the same search in a different browser without my SO login it show the same first three hits now from fifteen hours ago and has the banner bar "Welcome to Q&A ..." so it is seems unrelated to my account and is still reproducible here, albeit staler. – Rick Sladkey May 13 '11 at 1:15
Partly reproducible for me, sorting is correctly and the newest question is 36 minutes old...*but* the first and third question does have an answer. – Time Traveling Bobby May 13 '11 at 7:13
Can you still get this behavior? The 9 hours behind sounds very odd - we haven't had indexing issues in some time. The answers on the top few can happen, since we index every 15 minutes (30 at most - if there's a build in-between), and if an answer is added in the last 15 minutes, you may see that result. – Nick Craver Jun 13 '11 at 10:25
@Nick: This problem was confined to that day; I haven't seen it since. But judging from the upvotes, others some could repro it on that day. The tag [status-nonrepro] is fine. It was a real phenomenon for me but it was a transient problem. – Rick Sladkey Jun 13 '11 at 15:36
@Rick - I misread the dates, May 12th (I read yesterday) we did have a build issue with lucene on certain servers (app pools not actually dying), this was an issue then - which should have been resolved that night, making this [status-completed]. – Nick Craver Jun 13 '11 at 17:48
up vote 2 down vote accepted

There was a build-related issue on this day where some app domains just didn't want to die - even when the web sites were stopped in IIS, causing some indexing overlap which caused quite a few issues on certain web servers.

In our setup users are sticky mapped to a web server, so if you mapped to one having issues, you would have seen greatly lagged indexes, the 9 hours you see in the screenshot above is an example.

We resolved this and added a few more safeguards around so hopefully the prolonged overlapping app domains won't cause issues again (though they shouldn't have happened in the first place...)

share|improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .