>From what I have noticed , implementing your own
buffer manager, disk cache or using NIO or whatever
you want to call it, does improve performance in
"some" situations.
It avoids using the Operating System Disk Cache which
has some overheads.
In the database world it can sometimes yield higher
than 30% performance gains - assuming your index'es
are much larger than available ram.
This would make sense for a smaller implementation of
nutch , say, for intranet/enterprise search rather
than internet search.

Anyway, someone posted that a distributed filesystem
would be better and I guess that's the reason google
uses its own distributed filesystem.
Here is a paper that I'm sure someone else posted here
at some point.
http://www.cs.rochester.edu/sosp2003/papers/p125-ghemawat.pdf

-- Ab's

--- Byron Miller <[EMAIL PROTECTED]> wrote:
> Has anyone done any testing of converting
> read/writes
> & io to NIO or do you think that wouldn't offer much
> gain?
> 
> 
>
-------------------------------------------------------
> This SF.Net email is sponsored by The 2004
> JavaOne(SM) Conference
> Learn from the experts at JavaOne(SM), Sun's
> Worldwide Java Developer
> Conference, June 28 - July 1 at the Moscone Center
> in San Francisco, CA
> REGISTER AND SAVE! http://java.sun.com/javaone/sf
> Priority Code NWMGYKND
> _______________________________________________
> Nutch-developers mailing list
> [EMAIL PROTECTED]
>
https://lists.sourceforge.net/lists/listinfo/nutch-developers
> 



-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND
_______________________________________________
Nutch-developers mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/nutch-developers

Reply via email to