Herbert Poetzl wrote:


not that I'm aware of, but memory fragmentation
is a general problem (see lkml and Macelo working on
that stuff), so I'd suggest to try with a vanilla


I understand there's something wrong there with 2.6 kernels...

kernel (no vserver patches) and wait until the
order-4 allocation fails (which is probably after


In my case it's just order-0 allocations. What is strange is that it seems related
to the TCP treason uncloacked seen -which should not be the case as if I understand correctly, order-0 allocation error just means
a temporary failure ?-


a few hours of stress testing)



Rsync seems very good at that ...

For my part, testing without vserver will be very hard... This means relocating all others vservers & put the vserver programs
at the host level. Definitively not something I can do in some minutes.


anyway, I'll revisit the vserver code in the next
view days and will have a deep look for possible
connections ...

HTH,
Herbert



Could be a good idea.

In the same time I'll dig a bit further. For exemple I used this on the kernel config
Allocate 3rd-level pagetables from highmem -> Y


Maybe this has the exact reverse effect ?? I change this to No & boot another kernel in some minutes...

Sincerely yours,

--
Yann Dupont, Cri de l'universit� de Nantes
Tel: 02.51.12.53.91 - Fax: 02.51.12.58.60 - [EMAIL PROTECTED]

_______________________________________________
Vserver mailing list
[EMAIL PROTECTED]
http://list.linux-vserver.org/mailman/listinfo/vserver

Reply via email to