[c-nsp] 7206vxr reloading with software forced crash?

[Bruce Pinsky]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Francisco Rivas wrote:
> Hi all,
> 
> I've got a serious problem with one of my 7206vxr. It's reloading itself
> several times, with a "Software forced crash". I've seen this behavior
> before, and it won't happen too often, but this router in particular is
> reloading at least twice a day. It doesn't have too much traffic (just
> 30-40Mb), and the CPU load is < 50%. 
>  But, the main worry that I have about this, is that I've replaced the
> chassis ENTIRELY, and the new chassis is giving me the same error. Look:
> 
> 
> System was restarted by error - a Software forced crash, PC 0x606AED30
> at 14:47:49 UTC Wed Aug 31 2005
> 7200 Software (C7200-P-M), Version 12.2(14)S14, RELEASE SOFTWARE (fc1)
> Technical Support: http://www.cisco.com/techsupport
> Compiled Fri 08-Apr-05 13:48 by ccai
> Image text-base: 0x60008BD4, data-base: 0x612C8000
> 
> 
> This is the output of "sh region address 0x606AED30" on that chassis:
> 
> router#show region address 0x606AED30
> Address 0x606AED30 is located physically in :
> 
>   Name  : text
>   Class : IText
>   Media : R/O
>   Start : 0x60008BD4
>   End   : 0x612C7413
>   Size  : 0x012BE840
> 
> 
> 
> The old chassis was giving a similar error, not the same address, but it
> was on the IText class too.
> I've tried several IOS versions (12.2(14)S3, 12.2(14)S14, 12.2(18)S4)
> and all of them are crashing with a similar error. Always a software
> forced crash, always the same region address in the error (the region
> varies from IOS to IOS, but with the same IOS, is the same region in
> every crash), and always is a region on the IText class. And this is
> with 2 totally different chassis (both with NPE300 / 256Mb RAM / 6 port
> VXR Chassis).
> 
> 
> The last part of the crashdump, just before the %Software-forced reload,
> is this:
> 
> 
> Aug 31 14:47:46 UTC: %SYS-3-OVERRUN: Block overrun at 20014AB8 (red zone
> 1F860000)
> -Traceback= 60692A68 60695A04 606976D0 60697920
> Aug 31 14:47:46 UTC: %SYS-6-MTRACE: mallocfree: addr, pc
>  6770DFB4,6000011A 6770DB88,60C14048 6770DB88,60C0DED8 6770DB88,40000202
>  69C13E0C,60E22530 64C26254,60E27F30 65360AA0,60E37E9C 69C13E0C,60E1F41C
> Aug 31 14:47:46 UTC: %SYS-6-MTRACE: mallocfree: addr, pc
>  69C13E0C,3000000E 6770DFB4,6000011A 6770DB88,60C12734 6770DB88,60C0DED8
>  6770DB88,40000202 6770DFB4,6000011A 6770DB88,60C19E98 6601376C,60C25CA0
> Aug 31 14:47:46 UTC: %SYS-6-BLKINFO: Corrupted redzone blk 20014AB8,
> words 860, alloc 6063AAB8, InUse, dealloc 0, rfcnt 1
> -Traceback= 60690550 60692A7C 60695A04 606976D0 60697920
> Aug 31 14:47:46 UTC: %SYS-6-MEMDUMP: 0x20014AB8: 0xAB1234CD 0xFFFFFFFE
> 0x0 0x61584DD8
> Aug 31 14:47:46 UTC: %SYS-6-MEMDUMP: 0x20014AC8: 0x6063AAB8 0x20015198
> 0x200143EC 0x8000035C
> Aug 31 14:47:46 UTC: %SYS-6-MEMDUMP: 0x20014AD8: 0x1 0x0 0x0 0x0
> 
> 
> Anyone can give me a hand on this?? I've already filtered ALL the
> traffic going right to the IP address(es) of the router, but this is
> still happening. I have full crashlogs and "sh stacks" and "sh tech"
> available from the last crashes, so if anyone needs more info, I can
> provide that without problems.
> There are no changes on the config of the router since a couple of
> months, so I can't blame that... anyone have any idea about this?
> 
> 


The router crashed because it detected a memory corruption.  The traceback
decode indicates that an IS-IS update packet was being received at the
time, but that may not be related since it was the periodic memory checking
process called checkheaps that forced the crash.

If you have all the show stack, show tech, additional logging etc, you
should probably open a case to research this issue further.

- --
=========
bep

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (MingW32)

iD8DBQFDFemlE1XcgMgrtyYRAqEUAKDxZkNA+qE/klcISj+TzbAkk/VLDQCff8fL
Mdiy/H9zDUoyG6B/EY0A8hI=
=mrP2
-----END PGP SIGNATURE-----