[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

console server feedback concerning dhcp

I realized when thinking about how I was going to do things, that my current organization of the two console servers, Anduril and Octopussy suffers when they are the same vlan. I can't have two dhcp servers running because this would lead to one of them handing out the wrong ip in the current setup.

I thought about it and came up with a solution that, in the end, gives us more reliability should one console server fail.

I will run DHCPD on both servers with the same config. Since the console servers all have static IPs, it does not matter which server handles the information. I could also upgrade dhcpd and use the failover setup of a primary and backup.

I will make sure that on each console server, the information for the digis (ips and /dev devices) are unique. Then, on octopussy, the two files required for the console access will be stored in /etc/anduril/ (remote and dgrp.backing.store). The same will be true on anduril, in /etc/octopussy/ those two files will be stored. Should one of the servers fail or go down, the quick solution to get console access is:
cat /etc/<machine>/dgrp.backing.store >> /etc/dgrp.backing.store
cat /etc/<machine>/remote | sed 's/port/d2port/g' >> /etc/remote

This is assuming none of our names start with 'port'. I'm sure there's a better way to do it, but that's the quickest off the top of my head.

At that point, all you have to do is /etc/rc.d/init.d/dgrp_daemon restart and you will be able to access all of the consoles.

I'm looking for feedback on this approach. Does anyone see any issues with it, is it okay, does anyone have any better solutions or ideas.. etc.