LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

RE: 2 box LVS-DR + squid

To: "LinuxVirtualServer.org users mailing list." <lvs-users@xxxxxxxxxxxxxxxxxxxxxx>
Subject: RE: 2 box LVS-DR + squid
From: "James Masson" <jmasson@xxxxxxxxxx>
Date: Thu, 31 Mar 2005 17:40:44 +0200
> 
> 
> > I can connect via telnet to the individual squid servers OK, and if 
> > try to connect to the VIP it works fine, as long as the realserver 
> > chosen is the localserver. If ipvsadm chooses gb-squid2, I get a 
> > timeout.
> 
> it's not working properly
> 
> > Hearbeat and failover seem to work; if I shut down gb- squid1, 
> > gb-squid2 gets the VIP.
> >
> > I hope I've dealt with the ARP problem by doing:
> >
> > net.ipv4.conf.all.arp_ignore = 1
> > net.ipv4.conf.all.arp_announce = 2
> 
> did you check the values that the kernel has (use cat)?
> I haven't used arp_ignore. I thought you had to set it for 
> the particular NIC (eg eth0).
> 


The kernel values were set correctly - The values are set at boot time
via /etc/sysctl.conf on gb-squid1
I've now set it for that NIC (eth0) too...

I've re-read the localnode how-to, and added in the (gb-squid2)
realserver VIP on lo:110
I've also added to gb-squid2:
net.ipv4.conf.lo.arp_ignore = 1
net.ipv4.conf.lo.arp_announce = 2

It seems to be working OK for now, though it does seem biased towards
gb-squid2 (gb-squid1 does get used though...). Is there any easy way to
check whether I've squashed the arp problem?

thanks

James

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