[rancid] Solved - Re: Nexus expect looping

Dan_Mitton at YMP.GOV Dan_Mitton at YMP.GOV
Wed Oct 7 18:38:15 UTC 2009


I have the answer!!...

clogin tries to do a 'terminal length 0' to turn paging off.  Under NX-OS, 
that command is only available to users with the roles 'vdc-admin' and 
'network-admin'.  My rancid user was logging in and was only a 
'vdc-operator', so the command failed, so the expect script would hang at 
the first --More-- prompt, until it timed out, then it would loop.

Now on to all of the other nxrancid problems.




To:     Dan_Mitton at YMP.GOV
cc:     john heasley <heas at shrubbery.net>, rancid-discuss at shrubbery.net 
Subject:        Re: [rancid] Nexus expect looping
LSN: Not Relevant - Not Privileged
User Filed as: Excl/AdminMgmt-14-4/QA:N/A

Tue, Oct 06, 2009 at 03:27:10PM -0700, Dan_Mitton at YMP.GOV:
> Ok, some more info I just discovered...
> 
> Our whole configuration is 3 Nexus and about 20 misc. other Cisco IOS 
and 
> CatOS switches/routers.  Everything seems to be working except 1 of the 
> Nexus servers.  It is running a slightly old version of its software 
then 
> the other 2...
> 
> Broken:
> 
> Software
>   BIOS:      version 1.2.0
>   loader:    version N/A
>   kickstart: version 4.0(1a)N2(1)
>   system:    version 4.0(1a)N2(1)
>   BIOS compile time:       06/19/08
>   kickstart image file is: 
bootflash:/n5000-uk9-kickstart.4.0.1a.N2.1.bin
>   kickstart compile time:  2/25/2009 0:00:00 [02/25/2009 08:29:12]
>   system image file is:    bootflash:/n5000-uk9.4.0.1a.N2.1.bin
>   system compile time:     2/25/2009 0:00:00 [02/25/2009 08:56:57]
> 
> Working:
> 
> Software
>   BIOS:      version 1.2.0
>   loader:    version N/A
>   kickstart: version 4.1(3)N1(1)
>   system:    version 4.1(3)N1(1)
>   BIOS compile time:       06/19/08
>   kickstart image file is: bootflash:/n5000-uk9-kickstart.4.1.3.N1.1.bin
>   kickstart compile time:  7/23/2009 17:00:00 [07/24/2009 00:53:59]
>   system image file is:    bootflash:/n5000-uk9.4.1.3.N1.1.bin
>   system compile time:     7/23/2009 17:00:00 [07/24/2009 01:42:52]
> 
> Doing 'top' on the machine, I see 'expect' doing its normal thing for a 
> while and then waiting (for a long time), before looping to 100% cpu.

yes, iirc from the debugging, whatever happened before (dont really care)
would leave it with the device having closed the connection to the rancid
host.  at that point, expect should receive an EOF when trying to read
from the half-closed (and drained) pty/file descriptor, but it doesnt.

Either try rolling back to the suggested tcl/expect or report it to fbsd
ports maintainers.

> I added some debugging lines to nxrancid, to print out the $linecnt and 
$_ 
> and found...
> 
>         linecnt = 57, [vlan 200
> ]
>         linecnt = 58, [  name dan-test
> ]
>         linecnt = 59, [
> ]
>         linecnt = 60, [Error: TIMEOUT reached
> ]
>         linecnt = 61, [Hangup
> ]
> 
> The 'Hangup' is (I assume) from me 'kill -HUP <pid>' the expect process. 

> If I resize my terminal window from 60 lines, down to 50 lines, I then 
> get...
> 
>         linecnt = 48, [tacacs-server directed-request
> ]
>         linecnt = 49, [
> ]
>         linecnt = 50, [Error: TIMEOUT reached
> ]
>         linecnt = 51, [Hangup
> ]
> 
> So it seems like the 'pager' is involved some how.
> 
> 
> 
> 
> 
> 
> 
> To:     Dan_Mitton at YMP.GOV
> cc:     john heasley <heas at shrubbery.net>, rancid-discuss at shrubbery.net 
> Subject:        Re: [rancid] Nexus expect looping
> LSN: Not Relevant - Not Privileged
> User Filed as: Excl/AdminMgmt-14-4/QA:N/A
> 
> Tue, Oct 06, 2009 at 12:45:46PM -0700, Dan_Mitton at YMP.GOV:
> > pkg_info shows me the same version of expect, but a newer version of 
> > tcl...
> > 
> > pkg_info | egrep "expect|tcl"
> > expect-5.43.0_3     A sophisticated scripter based on tcl/tk
> > tcl-8.4.19_3,1      Tool Command Language
> > 
> > Do you think I should go back to the older version of tcl?
> 
> Please try it; I dont use fbsd regularly, but I know the combination
> below is working.
> 
> > Dan
> > 
> > 
> > 
> > 
> > To:     Dan_Mitton at YMP.GOV
> > cc:     rancid-discuss at shrubbery.net 
> > Subject:        Re: [rancid] Nexus expect looping
> > LSN: Not Relevant - Not Privileged
> > User Filed as: Excl/AdminMgmt-14-4/QA:N/A
> > 
> > Tue, Oct 06, 2009 at 10:28:01AM -0700, Dan_Mitton at YMP.GOV:
> > > John,
> > > 
> > > Thanks for your reply.  I'm running FreeBSD 7.2:
> > > 
> > > FreeBSD host.ymp.gov 7.2-RELEASE FreeBSD 7.2-RELEASE #0: Fri May  1 
> > > 07:18:07 UTC 2009 
> > > root at driscoll.cse.buffalo.edu:/usr/obj/usr/src/sys/GENERIC  amd64
> > > 
> > > expect verion:
> > > 
> > > expect version 5.43.0
> > 
> > Its most likely your expect/tcl set.  A friend (and others) has had
> > this problem a few times.  He recently updated his ports and the set
> > of tcl/expect that he got with that update has been working.  pkginfo
> > on that host gives me:
> > 
> > expect-5.43.0_3     A sophisticated scripter based on tcl/tk
> > tcl-8.3.5_8         Tool Command Language
> > 
> > > 
> > > 
> > > 
> > > To:     Dan_Mitton at YMP.GOV
> > > cc:     rancid-discuss at shrubbery.net 
> > > Subject:        Re: [rancid]  Nexus expect looping
> > > LSN: Not Relevant - Not Privileged
> > > User Filed as: Excl/AdminMgmt-14-4/QA:N/A
> > > 
> > > Tue, Oct 06, 2009 at 10:01:23AM -0700, Dan_Mitton at YMP.GOV:
> > > > I am trying to get Rancid to scan a Cisco Nexus 5000, but I'm 
having 
> 
> > > > problems.
> > > > 
> > > > First, it tried to run the command 'show version build-info all', 
> > which 
> > > > doesn't exist, so I commented it out of the nxrancid file.
> > > 
> > > that should not be necessary, it should just skip it.
> > > 
> > > > Second, it seems to get most of the way through (I think), but 
then 
> > the 
> > > > expect process goes to 100% CPU until it is killed.
> > > > 
> > > > The Nexus 5000 is running:
> > > > 
> > > > Software
> > > >   BIOS:      version 1.2.0
> > > >   loader:    version N/A
> > > >   kickstart: version 4.1(3)N1(1)
> > > >   system:    version 4.1(3)N1(1)
> > > >   BIOS compile time:       06/19/08
> > > >   kickstart image file is: 
> > bootflash:/n5000-uk9-kickstart.4.1.3.N1.1.bin
> > > >   kickstart compile time:  7/23/2009 17:00:00 [07/24/2009 
00:53:59]
> > > >   system image file is:    bootflash:/n5000-uk9.4.1.3.N1.1.bin
> > > >   system compile time:     7/23/2009 17:00:00 [07/24/2009 
01:42:52]
> > > > 
> > > > and has 1 - 2148 hung off of it.
> > > > 
> > > > Any ideas??
> > > 
> > > Most like its expect on you O/S.  What OS and expect version?  did 
you
> > > look at the o/s section on the rancid web page.
> > > 
> > > 
> > 
> > 
> 
> 


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.shrubbery.net/pipermail/rancid-discuss/attachments/20091007/7f16cb76/attachment.html 


More information about the Rancid-discuss mailing list