[rancid] Re: Infinate loop on clogin

[john heasley]

Fri, May 05, 2006 at 06:44:29PM +0200, Michael Stefaniuc:
> Sutherland, James wrote:
> > Interestingly we managed to accidentally get rancid set so that an
> > Efficient 5871 was set as a cisco.
> > 
> > What resulted was an infinite loop that causes rancid to hang forever
> > until the telnet to that device is killed.
> > 
> > I wasn't sure if this would be considered a bug, but I thought I'd send
> > it to this list as an example of this problem, which we've seen several
> > times.
> Call clogin with -t to give it a timeout. rancid uses clogin -t 90 to 
> prevent such loops being to disastrous.

That will do nothing, as the timeout only takes effect when there is no
input match.  Here, there's obviously a match.

> > 
> > Example:
> > Fri 8:58am {rancid at prometheus:[~/bin]} ./clogin 172.28.41.69
> > 172.28.41.69
> > spawn telnet 172.28.41.69
> > Trying 172.28.41.69...
> > Connected to 172.28.41.69.
> > Escape character is '^]'.
> > 
> > Efficient 5871 IDSL Router (5871-001/2) v5.3.80 Ready
> > Login: ******
> > Wrong password!   Try logging in again.
> > Login: ******

So its a bug/bad assumption that devices stop prompting and disconnect
after a few login failures.