On 11/6/2012 2:01 PM, Josh Turmel wrote:
Running Solr 3.3
We're running into issues where deleting individual documents (by ID) will
timeout but it only seems to happen when our hourly delta index is being
ran to pull in new documents, is there a way to work around this?
Are you experiencing really long commit times with your delta indexing,
or are you running an optimize after the delta index? That's the only
thing I can think of that might cause problems like that.
General recommendations: 1) Reduce commit times. Large autowarm counts
on the Solr caches will make commits very slow. Also, make sure you've
got enough RAM so your OS disk cache runs efficiently. 2) Don't
optimize often, unless the index is really really small and it can
complete quickly. On an index that never sees document deletes or
re-indexes, optimizing is completely unnecessary. Keep in mind that
re-indexing an existing document does an implicit delete first. Since
you are deleting documents, you'll want to do an optimize occasionally
to keep your index clean, but you should do it infrequently - once a day
or similar.
Another critical note about optimizing: No matter how often you do it,
you'll want to make sure that no other updates are happening on that
index while it is optimizing. Since I know this may be the obvious next
question: I am not aware of any way to query Solr and ask it if an
optimize is underway. You'll need to coordinate it yourself.
Thanks,
Shawn
