Aviran wrote: > My test index is pretty small size, about 250 documents and about 24 > fields in each document. The test is done by starting 10 threads that > repeat simple one word query (each thread query on a different word). > Neither range nor wildcard query is done. > I let the test run for about a minute and then I do a full thread dump to > see the stack trace. > I use a single searcher which never gets closed.
Thanks for providing these details. Benchmarking with such a small index will emphasize per-query-term overheads, like dictionary lookup. This is probably why you've seen such a large speedup when removing some thread contention from that area. However, folks with larger indexes are much less likely to encounter this thread contention, since relatively less of their time will be spent in this area. Is this benchmark typical of your application? Doug This is just a subset of the entire index. In the project that I'm working on we have ~1500 documents (this is a dynamic number, since documents are being added and deleted from the index every day). What do you consider a large index, I can build one just for testing purposes and see what the difference is. Aviran --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
