hello pat, Thank you very much for the quick response. I will check the same with sphinx group as well. I am trying to simplify the complex query structure. When i reduce max_match to :1000, i could save about 50-100ms.
regards madhu On Feb 17, 12:38 am, Pat Allan <[email protected]> wrote: > Hi Madhu > > I've never dealt with a Sphinx index of that size, so I'm not sure what the > best approach is. I recommend you ask this same question on the Sphinx Forum, > see if the people there can help.http://www.sphinxsearch.com/forum/ > > Also, you may want to look into the following > setting:http://www.sphinxsearch.com/docs/manual-0.9.9.html#conf-preopen > > It's also worth noting that queries get slower the more complex they are, and > your previous email was in a thread with a complex query you wanted... if > there's any way of simplifying that, then that should help speeds too. Having > an explicit max_matches much higher than 1000 also keeps Sphinx running a bit > slower. > > Hope this is somewhat helpful. > > Cheers > > -- > Pat > > On 17/02/2010, at 3:35 PM, madhu wrote: > > > Is there any performance tuning parameters for sphinx. I am using the > > following values in my sphinx.yml: > > mem_limit: 256M > > read_buffer: 2M > > Is there any other things i can tune to improve the performance ? > > > I have 10 million records and my sphinx index folder is about 2.5gb. > > and using "extended" search queries with max_matches: 2000. Its taking > > almost 1.5 seconds for one search query in sphinx. The average DB > > response time is 20 - 100ms. I want to reduce sphinx delay to 500ms or > > less. Is it possible ? > > > Other thing I observed is though the index size is 2.5gb, searchd only > > occupies 500mb of RAM when started. then when the queries are running, > > the memory usage goes between 500mb and 700mb. Is there a way to load > > entire index to memory so that i can completely eliminate the IO > > reads ?(I have enough RAM) > > > I am running sphinx server in one machine, DB in another machine and > > Web-Server on a third machine. > > > appreciate any help. > > > tx > > madhu > > > -- > > You received this message because you are subscribed to the Google Groups > > "Thinking Sphinx" group. > > To post to this group, send email to [email protected]. > > To unsubscribe from this group, send email to > > [email protected]. > > For more options, visit this group > > athttp://groups.google.com/group/thinking-sphinx?hl=en. -- You received this message because you are subscribed to the Google Groups "Thinking Sphinx" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/thinking-sphinx?hl=en.
