|
On Fri, 28 Apr 2000, Wayne wrote:
> >Packets coming from the RIP rather than the VIP is behind the identd
> >problem with sendmail and tcpwrappers (for telnet, ftp). What if
> >we put the VIP on (a non-arp'ing) eth0 on the realservers and put
> >the RIP on (an arp'ing) eth1. This would add about $20 to the cost of
> >each realserver, which isn't a big deal. Then we'd have to handle
> >affinity between identd and sendmail etc.
>
> Joe,
>
> We tried and it works! We put VIP on eth0, and RIP on eth0:1 in
> NAT mode and it works fine.
! :-)
I didn't realise you were in VS-NAT mode, as you don't have the VIP
in the realservers. I thought you must be in VS-DR.
So what has to come from the VIP to do whatever you need to do?
Just need to figure out how to do it
> during reboot, since this is done by playing with ifconfigure command.
> Once we swap them around, the going out IP address is the VIP
> address. But if LVS box reboot, you just have to redo it again.
Is the problem that you boot up with several machines all on the same IP?
Is the VIP on an arp'ing interface when you're done?
How about having 2 cards like I suggest above?
Or (haven't tested this) have eth0, eth0:1 both non-arping and
putting arp entries into the director with arp -f /etc/ethers
for each of the RIPs?
Joe
--
Joseph Mack mack@xxxxxxxxxxx
|
