|
Hi,
patrick edwards wrote:
>
> I can get my lvs works with no problem. However with in a matter
> of an hour or two my bandwidth drops to virtually nothing and the CPU load
> goes ballistic. I have a 100Mbit internal network, but at times i'm lucky
> to see 50Kps. Here is my LVS topology
>
> The director:
> DIP: xxx.xxx.214.3
> VIP: xxx.xxx.214.20 (webservers) and xxx.xxx.214.30 (sql)
> software:
> Mandrake 8.1
> kernel 2.4.8-26mdk
> ipvsadm 1.19-1mdk
> other notes:
> This box doubles as a bridging-firewall
>
> Real Servers:
> RIP1: xxx.xxx.214.21 and xxx.xxx.214.31 (RedHad 7.1)
> RIP2: xxx.xxx.214.22 and xxx.xxx.214.32 (Mandrake 8.1)
> Both are using the ipchain methd to avoid the ARP problem
> Both have dual eth0 eepro100s bonded
>
> Anyways like i said the cluster works perfectly. Then out of no where the
> bandwidth seemlingly disappears and the CPU rises dramatically. I dont
> know which comes first, but even ftping over the loopback gives god awful
> ftp speeds.
>
> Has anyone run across this? Or is there another possible reason this
> would occur? Perhaps upgrading the kernel on the director? Any help or
> tips would be more then welcome!
We had similar problems at our test servers for kernel 2.4
it turned out that ipchains under kernel 2.4 does full
connection-tracking and makes the
system slow. Try to use iptables or the arp-patch instead.
Cheers,
Chris
>
> _______________________________________________
> LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
> Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
> or go to http://www.in-addr.de/mailman/listinfo/lvs-users
|
