Now the question is from the previous replies you gave. I have moved the searches over to a dual processor Xeon 2.8 ghz machine with 3 gigs of ram. I have merged my segments and have right at 3 million pages. Doing single word searches most querys sit around .9 seconds. For multiple word querys most searches are around 1.5 seconds and longer.
That sounds slow to me. The index is on the local disk of the machine? You've run some other queries first to warm up the system before you benchmark? This is typical web content, i.e., not huge documents? Are you clustering results?
Also, check that the merged index is correctly found. The message "opening merged index in ..." should be printed. Do you see that?
I'd expect, with 3M pages on the hardware you describe, that you should be able to get more than 10 two-word searches/second throughput.
Doug
------------------------------------------------------- This SF.Net email is sponsored by BEA Weblogic Workshop FREE Java Enterprise J2EE developer tools! Get your free copy of BEA WebLogic Workshop 8.1 today. http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click _______________________________________________ Nutch-general mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/nutch-general
