|
Hi there,
i have a urgent problem here:
We've a well visited site (about 100 GB per day), PHP based with an
mySQL database. Because a single server was unable to handle the load,
I've installed a cluster (4 Linux based webservers) with two
loadbalancing servers on LVS base and own database server. It went to
public use some days ago.
Unfortunately, the web servers are freezing for some seconds from time
to time. The phenomenon occurs in unregular intervals (1-5 minutes)
indepentently on each server.
During these freezes, I can still type commands into the shell, but they
get not executed until the freeze is over. Afterwards, the load
increases rapidly, and some apache threads seem to die.
This makes me believe that something is wrong with my Apache and/or PHP.
I've tried some combinations of Apache and PHP on the webservers, but
the strange behaviour is still there. I rule out both hardware failure
and external attacks.
So I won't rule out a problem with the kernel, or the
loadbalancer, and that's the reason I've posted this question on this
list.
All systems are running on Debian 3.0r1 woody, updated to "testing". I'm
using LVS with direct routing through the internal network of my
cluster.
Has someone had similar pheonomenons? Or some ideas or hints?
Thank you in advance,
Jan
|
