The search index is updated every 15 minutes, so you can see some lag/mismatch here since the post data we pull as a result of searching our index is real-time.
Also yesterday especially our indexing was a little....wacky?...due to changing our core network switches (another update here). We index on our web tier for a few reasons:
- we have greatly under-utilized hardware there
- we're always happy to give the DB server even more breathing room
- it gives us a great deal of redundancy
Now, that being said, it doesn't really work well when yanking cables killing index sessions mid-index, which caused more delay than usual yesterday, the indexes got hours behind on Stack Overflow particularly because some directory locks couldn't be cleared correctly. This issue was fixed in a deploy last night and we kicked off an optimize to slim all those fragments back down as well, indexes are now updating normally.