On Fri, 2011-09-09 at 18:48 +0200, Mike Austin wrote: > Our index is very small with 100k documents and a light load at the moment. > If I wanted to use the smallest possible RAM on the server, how would I do > this and what are the issues?
The index size depends just as much on the size of the documents as the number, but assuming that your documents are relatively small, I don't see any issues. 100K is such a small amount that you will get fair OS caching even on a very low memory server. Plain searches works well with low memory, but faceting might be tricky as it requires a temporary memory overhead for each concurrent search. Limiting the number of concurrent searches to 1 or 2 might be a good idea. During tests of hierarchical faceting with Solr trunk, I tried running with 32MB for a very simple 1 million document index and it worked surprisingly well (better with 48MB though). For stable Solr I would expect the memory requirement to be somewhat higher. Could you tell us what you're aiming at? What is low memory for you, how large is your index in bytes, what response times do you hope for and what is the expected query rate?
