[rancid] IBM BNT Layer 2/3 Copper Gigabit Ethernet Switch no config collected

heasley heas at shrubbery.net
Fri Dec 16 17:44:03 UTC 2016


Tue, Dec 13, 2016 at 10:30:08AM +0300, Alexander R:
> On Tue, Dec 13, 2016 at 12:19 AM, heasley <heas at shrubbery.net> wrote:
> 
> > Mon, Dec 12, 2016 at 12:01:24PM +0300, Alexander R:
> > > my configuration:
> > > centos 7 x86_64
> > > i've issue in rancid 3.5.1 and 3.6.0
> >
> > thank you for including this.
> >
> > > i've issue to collect configs from switch modules "BNT Layer 2/3 Copper
> > > Gigabit Ethernet Switch Module for IBM BladeCenter (5.2.8 (FLASH
> > image1))",
> > > it has IP 192.168.1.11
> > >
> > > It's cisco-like CLI, so i've created my own "ibm" type in
> > rancid.types.conf:
> > >
> > > # cat /etc/rancid/rancid.types.conf
> > > ibm;script;rancid -t ibm
> > > ibm;login;clogin
> > > ibm;module;ios
> > > ibm;inloop;ios::inloop
> > > ibm;command;ios::WriteTerm;show running-config
> > >
> > > It's working fine with other switch module  "IBM Networking OS Virtual
> > > Fabric 10Gb Switch Module for IBM BladeCenter (7.8.3 (FLASH image1))" and
> > > it have IP 192.168.1.12, but not with "BNT Layer 2/3 Copper Gigabit
> > > Ethernet Switch Module for IBM BladeCenter (5.2.8 (FLASH image1))"
> > >
> > > Fail device:
> > > # rancid -t ibm -d 192.168.1.11
> > > loadtype: device type ibm
> > > loadtype: found device type ibm in /etc/rancid/rancid.types.conf
> > > executing clogin -t 90 -c"show running-config" 192.168.1.11
> > > PROMPT MATCH: sw11#
> > > HIT COMMAND:sw11#show running-config
> > >     In WriteTerm: sw11#show running-config
> > > 192.168.1.11: End of run not found
> > > end
> > >
> > > In 192.168.1.11.raw i see config issued by "show run..."
> > >
> > > Ok device:
> > > # rancid -t ibm -d 192.168.1.12
> > > loadtype: device type ibm
> > > loadtype: found device type ibm in /etc/rancid/rancid.types.conf
> > > executing clogin -t 90 -c"show running-config" 192.168.1.12
> > > PROMPT MATCH: sw12#
> > > HIT COMMAND:sw12#show running-config
> > >     In WriteTerm: sw12#show running-config
> > > 192.168.1.12: End of run not found
> > > end
> > >
> > > ps. 192.168.1.11 has NO words "exit" in config, because interface isn't
> > > configured (such a configuration plan).
> > > 192.168.1.12 has words "exit", because interfaces is configured.
> > >
> > > Could you help me?
> >
> > ios::WriteTerM() will expect to find and "end" marker in the config,
> > without
> > which it will not set $found_end.  if this device does not have this marker
> > or something similar that a customer version of ios::WriteTerm could use,
> > you can use a custom version of ios::WriteTerm that just counts lines,
> > which
> > ios::WriteTerm already does if the input is an ASA (or a few others), which
> > lack an end marker.
> >
> 
> Hi,
> thank you for responce.
> 
> It's my configs
> 
> Ok device:
> rancid]# tail -n5 192.168.1.12.raw
> ntp secondary-server 192.168.1.1
> !
> end
> 
> sw12#exitReceived disconnect from 192.168.1.12: 11: Logged out.
> rancid]#
> 
> Fail device:
> rancid]# tail -n5 192.168.1.11.raw
> ntp secondary-server 192.168.1.1
> !
> end
> 
> sw11#exitReceived disconnect from 192.168.1.11: 11: Logged out.
          ^^ this is likely the problem; the lousy thing didnt echo the
CR.  inloop() expects the CR and set $clean_run.  you'll have to adjust
the match; the lousy exos does this too.

> rancid]#
> 
> So, both switches have "end" word at the end of file. Any ideas? I don't
> know why it's not worked on same "text" files
> 
> 
> > the purpose of this is that some devices behave poorly in low memory
> > conditions, eg: memory leaks.  some appear to succeed to produce a config,
> > but output nothing.  requiring the marker prevents rancid from dropping
> > the config from the repo.
> >



More information about the Rancid-discuss mailing list