[rancid] Re: distributed RANCID setup

[John Adams]

You should be able to implement this pretty easily if you can get all of
your remote sites to access the CVS server. Instead of checking in the cvs
tree locally, you set CVSROOT To be the remote CVS server, and checkin/out
from there.

You can also run CVS over ssh with shared keys for additional security.

Read here for details on using CVS over SSH.

http://cvs.ecoinformatics.org/HOWTO-cvs-over-ssh.html

If you can't do this and you have to resync all of the repositories with the
master one by hand, it's going to involve quite a bit of scripting and work.

-john

On 10/17/06 9:13 AM, "William" <willay at gmail.com> wrote:

> Hi list,
> 
> Has anyone implemented RANCID where the collectors are distributed?
> 
> We currently have FreeBSD boxes dotted around sites for management
> purposes, with a central server that sits in our office and can
> communicate with each stub server. Currently our central server does
> all the RANCID/CVS work, however, there are some devices that we
> cannot reach for reasons out of my control.
> 
> What I would like to do is have each stub server access the devices
> within its reach, keeping its own repository then sync with the
> central whenever possible, meaning we can check our local (central)
> server to us to check configs but also not miss that critical config
> change when our links are down to our various stub servers.
> 
> Setups/configs/comments/papers/man pages/clues are all appreciated,
> 
> Thank you for your time.
> 
> Regards,
> 
> Will
> _______________________________________________
> Rancid-discuss mailing list
> Rancid-discuss at shrubbery.net
> http://www.shrubbery.net/mailman/listinfo.cgi/rancid-discuss

-- 
John Adams                         eLine
Sr. Network Engineer
jadams at eline.com                   537 Stevenson St
415-865-5139                       Suite 200
415-425-3551 (cell)                San Francisco, CA 94103