I'm assuming you're running the hobbit client on the host running the tunnel. On the server's hobbit-clients.cfg :

HOST=foo PROC="ssh -N -g -L 2022:127.0.0.1:22 -L 15294:10.20.2.11:15294 -R15294:10.0.2.11:15294 1.2.3.4"

This make sure that the process is running, but gives no assurance that all tunnel segments are ok ( I experience that problem often ). So you could also add :

PORT LOCAL=%[.:]15294 STATE=LISTEN "TEXT=foo bar"
PORT LOCAL=%[.:]2022 STATE=LISTEN "TEXT=foo bar 2"

This makes sure that "-L 2022:127.0.0.1:22" and "-L 15294:10.20.2.11:15294" are still up.

You can even add net tests for socket 2022 and 15294 by defining them in bb-services and having the hobbit server check them against that host.

Hope this helps.

Daniel Bourque Systems/Network Administrator Weather Data Inc

Office (316) 266-8013 Office (316) 265-9127 ext. 3013 Mobile (316) 640-1024

Hobbit User wrote:

On Fri, June 22, 2007 16:49, Lee J. Imber wrote:

...

Greet,

Is there a way to monitor a users specific process?

I have a ssh tunnel that needs to stay up. So how would I setup a test that monitors this:

user1 78512 0.0 0.5 3372 2416 p2- S 7:43PM 0:01.72 ssh - N -g -L 2022:127.0.0.1:22 -L 15294:10.20.2.11:15294 -R 15294:10.0.2.11:15294 1.2.3.4

The first local-to-remote (-L) mapping looks like a tunnel to an sshd on the other side, and could be checked with a client ext script to make sure the sshd is responding (if it's on the Hobbit server, should be testable with an ssh:2022 on localhost). The following pair of local and remote mappings look like a closed loop, and it's not clear to me how it might be used or tested--tcping perhaps?. Note that the latest versions of ssh/sshd support the use of a "bind" address in this syntax, so you could do -L 127.0.0.x:2022:127.0.0.1:22, for instance, to differentiate things by ip on the initiating end.

To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk