Ignite UX query

[Adam Noble]

Hello,

I'm hoping somebody can help. I am implementing an automated schedule whereby weekly ignite images are taken from all of our servers and stored on the central ignite server. I have simply set up a cron job on each client to push the images to the ignite server at a specific time. We have recently implemented a management LAN to segregate certain traffic. The problem I have is that I want to ensure that when I run a make_net_recovery from the client it uses the secondary LAN card to do this. Does anyone know how I ensure this is the case.

Thanks

[Ian Dennison_1]

Try using a routing entry in the routing table that ensures that the subnet you have set up has a route entry. It may also help to refer to the hosts via the Hostname that resolves to the secondary LAN.

route add net 94.22.44 netmask 255.255.255.0 94.22.22.1

Also, try editing /etc/rc.config.d/netconf, adding another route entry.

Share and Enjoy! Ian

[Michael Steele_2]

You can either start the make_net_recovery operation from the ignite server or from the client. If you pull from client to server then the router may always choose the primary ip of the server even if you specify the other ip. Clarify this.

If you can get past this router issue then use the server to call the 2nd ip address of the client. Make sure the two ip's are in different subnets.

If you use dynamic routing assign preferences and filter traffic into the 2nd ip.

http://docs.hp.com/cgi-bin/fsearch/framedisplay?top=/hpux/onlinedocs/B9106-90011/B9106-90011_top.html&con=/hpux/onlinedocs/B9106-90011/00/00/43-con.html&toc=/hpux/onlinedocs/B9106-90011/00/00/43-toc.html&searchterms=dynamic%7croute%7cgated&queryid=

Direct connect with a cross over cable to the 2nd nic. If you've got enough ports on the ignite server. There are quad port LAN cards, for instance.

[Steven E. Protter]

All you need to do is develop a little script that does a few things:

1) Checks connectivity to the server.
2) If there is no connectivity, adds the route.
3) Once successfully making sure there is connectivity, runs make_net_recovery -s IP_of_ignite_server

The key scripting technique is checking the outcome of the first two steps

ping ignite_server try two pings, wait 30 seconds
return_code=$?

if [ return_code ne 0 ]
then
add route
route_rc=$?
else
make_net_recovery
fi

Once the script works interactively, schedule it in cron

SEP

[Adam Noble]

Thanks guys you have given me a few ideas, I will see if I can implement them now.