|
Ok I'll need some time to work through some of these suggestions and report
results. I'll keep ya posted. thx for the suggsetions.
On 8/3/07, Graeme Fowler <graeme@xxxxxxxxxxx> wrote:
>
> On Fri, 2007-08-03 at 15:16 -0600, Dan Baughman wrote:
> > One, physically separate load balancer. (ubuntu 6.06). Two physically
> > separate real servers. This is all brand spanking new, high end intel
> > hardware. Running DR ( with the loopback adapater on the 2k3 servers).
>
> Aha... a thought comes to mind here.
>
> > The real servers are using the same physical interface to respond to the
> > load balanced requests as to the direct requests.
>
> OK.
>
> > The more hops out the more the issue presents itself. I am trying to by
> > pass the loadbalancer in the problem case to see if it is even the
> issue.
>
> PMTUD, aka Path MTU discovery.
>
> Somewhere in the path between you and your clients is a router which
> isn't honouring ICMP replies telling it that DEST_UNREACH/FRAG_NEEDED.
>
> Try tuning your MTU down on both the W2K3 servers *and* the director so
> they're all 400 (you'll have to look up how to do this for your distro,
> but it usually isn't hard). It'll make the network seem less efficient
> on the face of it, but "less efficient" might mean "working", and if it
> does then you're sorted.
>
> If that fixes things, turn it up in increments of 100 bytes until it
> breaks; that way you find the "weakest link" and can then set the MTU to
> be lower than that value.
>
> Graeme
>
>
> _______________________________________________
> LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
> Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
> or go to http://lists.graemef.net/mailman/listinfo/lvs-users
>
|
