|
Hello,
On Tue, 21 Aug 2001, Nate Stone wrote:
> Hi all! I'm pretty new to LVS, but we've had some short term success and
> look forward to using the system as our site expands. However, we're
> having some problems I've never seen before and I really can't tell if LVS
> is causing them or not.
>
> We currently have 1 director and 2 real servers which use NFS to mount the
> same web content. We serve about 45 million pages a month and are now
> only on our 8th day of using LVS on our new server cluster. As we were
> ramping up traffic, we started to have some problems with the load average
> on our real servers going extremely high and then the Apache processes
> going into "D" mode (uninterruptible sleep) mode. This causes many
> problems which ultimately require us to reboot the affected real server.
It seems you have problems with your real server OS. May be
you can tell us the exact versions of the real server (if Linux) and
of the Apache. But that seems as incorrect process state management
in the kernel. What shows vmstat on the real server? Full CPU usage?
Any additional patches applied on the real servers?
> My first instinct is that we're saturating NFS, but I wanted to know if
> anyone thought that IPVS might be causing the problems or if anyone had
> any ideas that could lead to a solution.
>
> I hope it isn't an inappropriate question for this board - I'm just a
> little too new to LVS to be able to tell if it might be contributing to
> the problem so I was hoping that someone with a little more expertise
> might be able to definitively rule an LVS problem out.
Your real server should be ready for everything: attacks,
incorrect packets, etc. IMO, you can achieve the same D states even
without using LVS. The problem can't be in the director.
> Thanks a bunch.
>
> -Nate
> Nate Stone
> CTO
> keenspot.com
Regards
--
Julian Anastasov <ja@xxxxxx>
|
