Solr Timeouts during query aggregation

[Peter Lee <Pe...@proquest.com>]

Greetings,

I'm having a hard time locating information about how solr handles timeouts (if it DOES handle them) during a search across multiple shards in a solr cloud configuration.

I have found information and have confirmed through empirical testing as to how solr handles timeouts on each shard when a query is made to a collection. What I have NOT been able to find is information or settings related to the time it takes Solr to aggregate the results returned from multiple shards before returning the response to the user. Does Solr not have any sort of timeout on this operation?

For clarity, I'll repeat my question and try to explain it in more detail.

If I send a query to a solr cloud setup that has 6 shards, the query will be sent to each of the 6 shards that will each return some number of hits. The answers from each of the shards is sent back to the server that originally caught the query, and that original server must then aggregate the data from all of the different shards to produce a single set of hits to return to the user. I see how to use "timeAllowed" to limit the time of the search on each shard...but I was wondering if there was a separate timeout for the "aggregation" step just before the response is returned.

I am asking this question because our existing search technology has this behavior and setting, and I and I am trying to determine if there is a related feature within the solr technology. At this point, since I have not seen any documentation nor configuration settings for this feature, I am ready to take it as truth that Solr does NOT include this functionality. However, I thought I should ask the mailing list to see if I've missed something.

Thank you!

Peter S. Lee, Software Engineer Lead
ProQuest | 789 E. Eisenhower Parkway | Ann Arbor, MI, 48106-1346 USA
www.proquest.com