Multiple interfaces

[Phil Dibowitz]

On Fri, Jul 30, 2004 at 04:12:07PM -0700, PGuth at corp.terralycos.com wrote:
> You *could* do this by defining each interface as a service under a single 
> host.  For the check_command for each service you would have something 
> like check_ping_by_ip!<interface IP>.  That would give you the cleaner 
> host presentation.  But you still define one thing for each interface. And 
> you'd have to define check_ping_by_ip, of course.  If you actually need to 
> monitor multiple services for *each* interface it's going to be pretty big 
> and unwieldy.

Okay, lets take HTTP. It seems like HTTP would be in a "warning" state it it
doesn't respond on all interfaces, "fail" if it doesn't work on any, and
"okay" if it works on all.

NOC can go "HTTP is in a warning state on HOST"

You bring up a good point that a sysadmin can't look at Nagios and see what
interface is down, but sysadmins won't be looking at nagios, the NOC staff
will. And some of the NOC staff is less then stellar... and more than "it
works" or "it doesn't work" is a hard concept to get across. If HTTP dies,
getting them to understand that almaak0 through almaak5 is all the same host
is non-trivial, especially when you have hosts named like hpc001,
hpc002...hpc978.

Which brings us to the "hostgroups that represent hosts" option...

But yes -- services should be monitored on all interfaces... in my opinion...

-- 
Phil Dibowitz
Systems Architect and Administrator
Enterprise Infrastructure / ISD / USC
UCC 174 - 213-821-5427

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <https://www.monitoring-lists.org/archive/users/attachments/20040730/56e63f79/attachment.sig>