[Vacm-general] sometimes, 'unable to ping nexxus'

[Tony R. <tr...@lu...>]

Hi,

We are using VACM and Vash in our server farm of 15 or
so front end servers.  I must say that this is an incredible
tool and has saved us many trips and/or phone calls to the
co-lo facility!!! Thanks to the developers!!!

Here is the problem that I am trying to solve...

We have 2 machines that each have a rocketport, there is a nexxus
running on each of these and the nodes are split between these two
nexxi(?).  Another machine, 'master' it's own binary of Vash to
connect to each nexxus.  We want to be able to script some vacm
actions however vacm seems to not always respond, therefore we can
not rely on it for scripting just yet.  Here is the  error that we get
when we 'ping' each host with a nodestatus.

vash: Unable to ping nexxus at 192.168.12.15
vash: Unable to find connected Nexxus 192.168.12.15

and then, sometime later (usually immediately) it will work...
EMP:445:JOB_STARTED
EMP:445:NODESTATUS:/dev/ttyR16:DETECTED
EMP:445:JOB_COMPLETED

this happens intermittently and does not seem to favor
certain nodes over others. If I do this on all 26 of our 
nodes there will almost always be 1-3 nodes that respond 
in this way.

the 2 nexxus machines are Debian:
Linux mon-2 2.2.18 #1 SMP Wed Mar 14 11:48:40 PST 2001 i686 unknown
Linux mon-1 2.2.18 #1 SMP Wed Jun 20 12:06:32 PDT 2001 i686 unknown

and the master machine is Debian:
Linux master 2.2.17 #1 SMP Tue Dec 19 16:07:50 PST 2000 i686 unknown

I recently upgraded the version of VACM to 2.0.5 and it seems that
it helped some although it is still somewhat unstable.

Any ideas on what could be causing this?

Thanks!

-Tony





-- 
$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ 
$		Tony Rose                         $
$      Senior Web Systems Administrator           $
$     650-616-3911 - tr...@lu...          $
$   Free! Play for $1 Million every day!!!!       $
$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $