java client locking issue

[Dustin Sallings]

On Apr 4, 2007, at 15:35 , Steven Grimm wrote:

> If you're connecting to a remote host, network round-trips will  
> dominate the response time. You will definitely see better  
> throughput with multiple connections, especially if you're  
> requesting small objects. Even on the local host you might see a  
> marginal benefit.

	I measured a considerable improvement in the initial version of my  
implementation with a single thread over the multithreaded  
implementation.  I'll go ahead and try adding multiple connections  
per destination and see if it makes a difference.

> Also, there is a multithreaded version of memcached (in Subversion  
> only for now -- see the "multithreaded" branch -- though at some  
> point I hope it will become the main version of the code.) We are  
> successfully using it on a network of 4-core memcached hosts.

	Does that really give a considerable performance benefit over  
running four single-threaded processes on the same box?  I'd be  
concerned about locking (and correctness) on the server-side.

	Is there any documentation on how this branch uses threads?  It'd be  
an interesting read.

-- 
Dustin Sallings


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.danga.com/pipermail/memcached/attachments/20070404/02c89c1a/attachment-0001.htm