By not generating MD5 hashes on all words of indexed texts, processing
time is reduced by 30 to 50% on indexed documents with more than 1Mbytes
of plain text.
jdk-based logger tend to block
at java.util.logging.Logger.log(Logger.java:476) in concurrent
environments. This makes logging a main performance issue. To overcome
this problem, this is a add-on to jdk logging to put log entries on a
concurrent message queue and log the messages one by one using a
separate process.
- FTPClient uses the concurrent logging instead of the log4j logger
- some restructuring of the document counting and logging structures was necessary
- better abstraction of CrawlProfiles
- added deletion of logs to the index deletion option (if the index is deleted using the servlets) which is necessary to reset the domain counters for the page limitation
- more refactoring to get the LibraryProvider more clean
- some refactoring of the Condenser class
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7478 6c8d7289-2bf4-0310-a012-ef5d649a1542
this makes YaCy search results VERY fast for all verify=false search cases
and it enhances the search speed also for all other snippet-fetch cases.
With this change my peer performed 100 Queries Per Second (!!!) while doing 10 queries simultanously (!!!)
in an intranet index of 20000 URLs on my 16-core Mac
Check this yourself by doing:
cd bin
./searchtestmulti.sh
after finishing the run, divide 1000 by the given time per query (which is the qps for one thread)
and then multiply again by 10 (because 10 search threads has been started)
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7231 6c8d7289-2bf4-0310-a012-ef5d649a1542
