LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

Re: [lvs-users] DR stops working when server = director

To: lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Subject: Re: [lvs-users] DR stops working when server = director
From: Charlie Allom <charlie@xxxxxxxxxxxxxx>
Date: Mon, 29 Nov 2010 21:35:56 +0000
Hello,

I can confirm that these issues were exasperated by Ubuntu's since-lucid
default of turning on rp_filter.

this fixes:
for i in /proc/sys/net/ipv4/conf/*/rp_filter ; do echo "setting $i to 0" ; echo 
0 > $i ; done


Regards,
  C.

On Wed, Nov 03, 2010 at 08:15:08PM +0000, Charlie Allom 
<charlie@xxxxxxxxxxxxxx> wrote:
> Hello,
> 
> I am having an odd issue with a DR setup. When the director is running,
> and the VIP is on eth0:0, I can't reach the VIP over TCP. I know it's to
> do with LVS, because as soon as I swap the director by failing the
> heartbeat, the new director (app1) then fails in the same way..
> and the failing one (app0) then starts working!
> 
> a diagram:
> 
>         app0 (ipvsadm 1:1.25.clean-1)
>             eth0:0 94.228.67.2/32
>             eth0   94.228.67.7/28
> 
>         app1 (heartbeat backup)
>             lo:0   94.228.67.2/32
>             eth0   94.228.67.8/28
> 
>         app2 (plain old server)
>             lo:0   94.228.67.2/32
>             eth0   94.228.67.9/28
> 
> and here is the problem:
> 
>         20:03 app0:~% sudo /etc/init.d/nginx restart
>         Restarting nginx: nginx.
> 
>         20:04 app0:~% telnet 94.228.67.2 80
>         Trying 94.228.67.2...Connected to 94.228.67.2.
>         Escape character is '^]'.
>         ^]
>         telnet> Connection closed.
> 
>         20:04 app0:~% telnet 94.228.67.2 80
>         Trying 94.228.67.2...
>         telnet: Unable to connect to remote host: Connection timed out
> 
>         20:04 app0:~% sudo ethtool -k eth0
>         Offload parameters for eth0:
>         rx-checksumming: on
>         tx-checksumming: on
>         scatter-gather: off
>         tcp-segmentation-offload: off
>         udp-fragmentation-offload: off
>         generic-segmentation-offload: on
>         generic-receive-offload: off
>         large-receive-offload: off
> 
>         20:04 app0:~% sudo /etc/init.d/nginx restart
>         Restarting nginx: nginx.
> 
>         20:05 app0:~% telnet 94.228.67.2 80         
>         Trying 94.228.67.2...
>         telnet: Unable to connect to remote host: Connection timed out
> 
>         20:05 app0:~% sudo sysctl -p
>         net.ipv4.conf.lo.arp_ignore = 1
>         net.ipv4.conf.lo.arp_announce = 2
>         net.ipv4.conf.all.arp_ignore = 1
>         net.ipv4.conf.all.arp_announce = 2
> 
>         20:07 app0:~% uname -a
>         Linux app0 2.6.32-25-server #45-Ubuntu SMP Sat Oct 16 20:06:58
>         UTC 2010 x86_64 GNU/Linux
> 
> Strangely though, ICMP works throughout the TCP failing. I never lose a
> single ping.
> 
> These NICs are Broadcom NetXtreme II BCM5709's.
> 
> I have successfully setup this configuration on Debian lenny,
> 2.6.26-2-amd64, so I know it works.. I'm just hoping someone has seen
> this before and can point me in the right direction.
> 
> Regards,
>   C.
> -- 
>  020 7729 4797
>  http://playlouder.com/
> 
> _______________________________________________
> Please read the documentation before posting - it's available at:
> http://www.linuxvirtualserver.org/
> 
> 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

-- 
 020 7729 4797
 http://playlouder.com/

_______________________________________________
Please read the documentation before posting - it's available at:
http://www.linuxvirtualserver.org/

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

<Prev in Thread] Current Thread [Next in Thread>