|
Hi,
We're seeing an unusual problem which we're having great difficulty trouble
shooting. We have LVS load balancing JMS connections between a number of SUN
eXchange Processing domains and two Routers (SUN stcms IQ managers). During
testing, our testers see a slow down in the Processor to Router
communication. We've observed the output of ipvsadm -lc during the tests -
behaviour is normal and connections are load balanced evenly.
When we take the load balancers out of the mix, everything functions OK
(except for the scalability problem induced), so it appears there's a issue
with the load balancers but we have no clue as to what that issue maybe.
I hoped that there might be others out there who run JMS in a high
performance environment across LVS who may have seen similar symptoms and
could pass on the benefit of their experience.
I realise we're grasping at straws here, but that's a reflection of where
we're at! I apologise for the lack of information I can give you to analyse.
Cheers,
Bruce
--
m: +64 (0)21 962 085
h: +64 (0)4 384 6482
e: thompson.bj@xxxxxxxxx
skype: bjthompson
|
