[rancid] (no subject)

[Alan McKinnon]

On 02/12/2015 13:31, Harsha S Aryan wrote:
> hi,
> 
> i have some doubts in rancid can any one explane what is the below
> things status for
> 
> root@[server]:/# vi /var/lib/rancid/switches/router.db
> switch1.your-domainname.xyz:hp:up
> switch2.your-domainname.xyz:hp:up
> 1.2.3.4:hp:up
> 1.2.3.5:hp:down


This is all in the man pages somewhere, but here goes:

That's a config file from the rancid 2.3 series; the 3.x series uses a
semicolon ";" as a field separator and not a colon ":" as was used in
2.3. This is to avoid issues with IPv6 addresses; otherwise the format
is the same.

Each line has 3 fields:

device name: this can be it's fqdn or an IP address
device type: a device type as defined in rancid (see [1] below)
status: "up" means the device is working and rancid will poll it. Any
other value means down and rancid will not poll it

[1] A device type of "cisco" or "hp" does not have to mean the device
was made by Cisco or Hewlett Packard. If Acme Inc made a device that
copies Cisco IOS config closely, then the cisco type would probably work
for it. On the other hand, recent Cisco OSes (eg GSR, Nexus and XR) are
a lot different from classic IOS so the cisco type doesn't work for
them, you use types like cisco-gsr, cisco-nx and cisco-xr.

HP kit (liek Dell) has a reputation for being badge engineered and
sometimes it feels like every model is completely different. Usually you
have to find out what the actual OEM device is, and use that type.

Confusing? you betcha.

[2] There a possible optional fourth field in that file, it is treated
as a comment. I used it a lot to document the date and reason I marked
something down, and it shows up in the daily status emails. Useful tip!


-- 
Alan McKinnon
alan.mckinnon at gmail.com