I'm using resin to load-balance requests to two srun's on another 
server.  We're long-time Resin users and much experience with this 
particular version.  We've implemented load-balancing a number of times 
so are familiar with same.

In this particular scenario, I've observed while stress-testing the 
applications that requests are not being equally distributed to both 
services.  I've seen as many as 98 sessions on the first srun with the 
heap nearly exhausted with only 15 being served by the second and it's 
virtually idle.

At other times, I've seen the load more equally distributed.  If it's 
truly round-robin, how could this gross imbalance occur?  The behavior 
seems to be intermittent but when it's bad, it's really bad.

Thanks.

-- 
Scott Dudley
Senior Developer

Telesoft Corp. | 3443 N. Central Ave., 18th Floor | Phoenix, AZ, 85012

*o:* (602) 308-1115* *| *f:* (602) 308-1300 | *w:* www.telesoft.com 
<http://www.telesoft.com/?utm_source=signature&utm_medium=email&utm_campaign=signature>



_______________________________________________
resin-interest mailing list
resin-interest@caucho.com
http://maillist.caucho.com/mailman/listinfo/resin-interest

Reply via email to