Colossus – Guardian Failover Procedure
Now that c has become a file server this procedure is rather complicated. This hasn’t actually been tried, but these are the basic steps that should be followed:
If guardian dies we recreate it on colossus. Power down both computers, take the bottom two internal harddrives (the boot drives) out of c and store them safely. Then remove all the drives from g and replace them in the corresponding slots in c. Then boot. At this point different things could/should be done.
At this point g is back up but c, and all the files served on it are down. Here is what I’d do
2. c3t0d0 <SEAGATE-ST173404FCV-4303 cyl 14087 alt 2 hd 24 sec 424>
3. c3t1d0 <SEAGATE-ST173404FCV-4303 cyl 14087 alt 2 hd 24 sec 424>
4. c3t2d0 <SEAGATE-ST173404FCV-4303 cyl 14087 alt 2 hd 24 sec 424>
lo0: flags=849<UP,LOOPBACK,RUNNING,MULTICAST> mtu 8232
inet 127.0.0.1 netmask ff000000
ge0: flags=863<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST> mtu 1500
inet 220.127.116.11 netmask ffffff00 broadcast 18.104.22.168
ge0:1: flags=842<BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 22.214.171.124 netmask ffffff00 broadcast 126.96.36.199
at which point you can do an ifconfig ge0:1 up. Also, in /etc/rc3.d there is a fo_S99ifcon that could be renamed to S99ifcon to automatically bring up the interface at boot time.
If colossus dies the main thing will be to move the Fibre Channel disks to g. This basically will involve moving the FC cards from c to g then following steps 1-4 above, of course ge0:1 has the IP address 188.8.131.52 on colossus. There is also a spare FC card on the third shelf next to my desk, the self with all the TrUX documentation.
As always see me with any questions or comments – James Last updated 7/3/01