Discussion:
[rancid] Rancid not collecting for ExtremeXOS switches
Michael Tiarnaigh
2010-11-17 11:52:20 UTC
Permalink
Hi all.

My apologies if this has been asked before, but I have not as yet found
an answer for it if so. I have seen some others with similar issues from
trawling through Google searches, but as yet can't find a fix......





I have a bunch of new Extreme XOS switches (250's & 450's) which we're
currently adding to our network, and simply cannot get Rancid to poll
for them successfully:



TowercomTuam.1 # show version

Switch : 800216-00-01 0950G-80375 Rev 1.0 BootROM: 1.0.3.5 IMG:
12.3.3.6



Image : ExtremeXOS version 12.3.3.6 v1233b6 by release-manager

on Fri Oct 16 20:01:57 PDT 2009

BootROM : 1.0.3.5



We also have a few hundred of the older Summit-48si's which work just
fine. The issues I'm seeing with the new ExtremeXos switches are either
that I get constant "end of run not found" errors, or the Expect process
hangs indefinitely and consumes more and more resources on the server.



I have played around with my .cloginrc file and have added the following
entries for one switch:



add autoenable tctuam-switch.internal.nw 1

add method tctuam-switch.internal.nw ssh

add passprompt tctuam-switch.internal.nw "rancidadmin:"

add user tctuam-switch.internal.nw rancidadmin

add password tctuam-switch.internal.nw ******



I added the add passprompt tctuam-switch.internal.nw "rancidadmin:"
part, as it was hanging before this. When I run this through clogin from
the command line, it does appear to connect to the switch as normal and
I can then interact with it as expected:





[***@cron /usr/local/libexec/rancid]$ ./clogin
tctuam-switch.internal.nw

tctuam-switch.internal.nw

spawn ssh -c 3des -x -l rancidadmin tctuam-switch.internal.nw

Keyboard-interactive authentication

Enter password for rancidadmin:

ExtremeXOS

Copyright (C) 2000-2009 Extreme Networks. All rights reserved.

Protected by US Patent Nos: 6,678,248; 6,104,700; 6,766,482; 6,618,388;
6,034,957; 6,859,438; 6,912,592; 6,954,436; 6,977,891; 6,980,550;
6,981,174; 7,003,705; 7,017,082; 7,046,665; 7,126,923; 7,142,509;
7,149,217; 7,152,124; 7,154,861; 7,245,619; 7,245,629; 7,269,135.

========================================================================
======



Press the <tab> or '?' key at any time for completions.

Remember to save your configuration changes.



TowercomTuam.1 #

TowercomTuam.1 # show version

Switch : 800216-00-01 0950G-80375 Rev 1.0 BootROM: 1.0.3.5 IMG:
12.3.3.6



Image : ExtremeXOS version 12.3.3.6 v1233b6 by release-manager

on Fri Oct 16 20:01:57 PDT 2009

BootROM : 1.0.3.5

TowercomTuam.2 # q



Connection to tctuam-switch.internal.nw closed.





The issue is however, getting this to run via the rancid-run scripts.
Any time I enable this switch to collect via my router.db file, Rancid
seems to crumble!





Does anyone have any ideas on what I'm missing here? I'm sure others
have been through this before and have hit the same frustrations as
myself.



Thanks in advance for any help offered,



Michael.
john heasley
2010-11-17 22:17:19 UTC
Permalink
Post by Michael Tiarnaigh
I have a bunch of new Extreme XOS switches (250's & 450's) which we're
currently adding to our network, and simply cannot get Rancid to poll
which version of rancid on what O/S with which version of expect?
Post by Michael Tiarnaigh
TowercomTuam.1 # show version
12.3.3.6
Image : ExtremeXOS version 12.3.3.6 v1233b6 by release-manager
on Fri Oct 16 20:01:57 PDT 2009
BootROM : 1.0.3.5
We also have a few hundred of the older Summit-48si's which work just
fine. The issues I'm seeing with the new ExtremeXos switches are either
that I get constant "end of run not found" errors, or the Expect process
hangs indefinitely and consumes more and more resources on the server.
I have played around with my .cloginrc file and have added the following
add autoenable tctuam-switch.internal.nw 1
add method tctuam-switch.internal.nw ssh
add passprompt tctuam-switch.internal.nw "rancidadmin:"
add user tctuam-switch.internal.nw rancidadmin
add password tctuam-switch.internal.nw ******
I added the add passprompt tctuam-switch.internal.nw "rancidadmin:"
part, as it was hanging before this. When I run this through clogin from
the command line, it does appear to connect to the switch as normal and
tctuam-switch.internal.nw
tctuam-switch.internal.nw
spawn ssh -c 3des -x -l rancidadmin tctuam-switch.internal.nw
Keyboard-interactive authentication
ExtremeXOS
Copyright (C) 2000-2009 Extreme Networks. All rights reserved.
Protected by US Patent Nos: 6,678,248; 6,104,700; 6,766,482; 6,618,388;
6,034,957; 6,859,438; 6,912,592; 6,954,436; 6,977,891; 6,980,550;
6,981,174; 7,003,705; 7,017,082; 7,046,665; 7,126,923; 7,142,509;
7,149,217; 7,152,124; 7,154,861; 7,245,619; 7,245,629; 7,269,135.
========================================================================
======
Press the <tab> or '?' key at any time for completions.
Remember to save your configuration changes.
TowercomTuam.1 #
TowercomTuam.1 # show version
12.3.3.6
Image : ExtremeXOS version 12.3.3.6 v1233b6 by release-manager
on Fri Oct 16 20:01:57 PDT 2009
BootROM : 1.0.3.5
TowercomTuam.2 # q
Connection to tctuam-switch.internal.nw closed.
The issue is however, getting this to run via the rancid-run scripts.
Any time I enable this switch to collect via my router.db file, Rancid
seems to crumble!
Does anyone have any ideas on what I'm missing here? I'm sure others
have been through this before and have hit the same frustrations as
myself.
Thanks in advance for any help offered,
Michael.
_______________________________________________
Rancid-discuss mailing list
http://www.shrubbery.net/mailman/listinfo.cgi/rancid-discuss
Michael Tiarnaigh
2010-11-18 10:00:12 UTC
Permalink
Sorry, I should have included this with my original mail.

Rancid is running on FreeBSD 7.0-STABLE with what appears to be v2.3.2
and expect v5.43.0

Thanks,
M



-----Original Message-----
From: john heasley [mailto:***@shrubbery.net]
Sent: 17 November 2010 22:17
To: Michael Tiarnaigh
Cc: rancid-***@shrubbery.net
Subject: Re: [rancid] Rancid not collecting for ExtremeXOS switches
Post by Michael Tiarnaigh
I have a bunch of new Extreme XOS switches (250's & 450's) which we're
currently adding to our network, and simply cannot get Rancid to poll
which version of rancid on what O/S with which version of expect?
Post by Michael Tiarnaigh
TowercomTuam.1 # show version
Switch : 800216-00-01 0950G-80375 Rev 1.0 BootROM: 1.0.3.5
12.3.3.6
Image : ExtremeXOS version 12.3.3.6 v1233b6 by release-manager
on Fri Oct 16 20:01:57 PDT 2009
BootROM : 1.0.3.5
We also have a few hundred of the older Summit-48si's which work just
fine. The issues I'm seeing with the new ExtremeXos switches are either
that I get constant "end of run not found" errors, or the Expect process
hangs indefinitely and consumes more and more resources on the server.
I have played around with my .cloginrc file and have added the
following
Post by Michael Tiarnaigh
add autoenable tctuam-switch.internal.nw 1
add method tctuam-switch.internal.nw ssh
add passprompt tctuam-switch.internal.nw "rancidadmin:"
add user tctuam-switch.internal.nw rancidadmin
add password tctuam-switch.internal.nw ******
I added the add passprompt tctuam-switch.internal.nw "rancidadmin:"
part, as it was hanging before this. When I run this through clogin from
the command line, it does appear to connect to the switch as normal and
tctuam-switch.internal.nw
tctuam-switch.internal.nw
spawn ssh -c 3des -x -l rancidadmin tctuam-switch.internal.nw
Keyboard-interactive authentication
ExtremeXOS
Copyright (C) 2000-2009 Extreme Networks. All rights reserved.
Protected by US Patent Nos: 6,678,248; 6,104,700; 6,766,482;
6,618,388;
Post by Michael Tiarnaigh
6,034,957; 6,859,438; 6,912,592; 6,954,436; 6,977,891; 6,980,550;
6,981,174; 7,003,705; 7,017,082; 7,046,665; 7,126,923; 7,142,509;
7,149,217; 7,152,124; 7,154,861; 7,245,619; 7,245,629; 7,269,135.
========================================================================
Post by Michael Tiarnaigh
======
Press the <tab> or '?' key at any time for completions.
Remember to save your configuration changes.
TowercomTuam.1 #
TowercomTuam.1 # show version
Switch : 800216-00-01 0950G-80375 Rev 1.0 BootROM: 1.0.3.5
12.3.3.6
Image : ExtremeXOS version 12.3.3.6 v1233b6 by release-manager
on Fri Oct 16 20:01:57 PDT 2009
BootROM : 1.0.3.5
TowercomTuam.2 # q
Connection to tctuam-switch.internal.nw closed.
The issue is however, getting this to run via the rancid-run scripts.
Any time I enable this switch to collect via my router.db file, Rancid
seems to crumble!
Does anyone have any ideas on what I'm missing here? I'm sure others
have been through this before and have hit the same frustrations as
myself.
Thanks in advance for any help offered,
Michael.
_______________________________________________
Rancid-discuss mailing list
http://www.shrubbery.net/mailman/listinfo.cgi/rancid-discuss
Arjan Oosting
2010-12-23 10:37:36 UTC
Permalink
Hi Michael,

I use the attached patch to get RANCID working with out ExtremeXOS
switches.

Greetings Arjan

Loading...