[afnog] Cisco AS5300 issues

Mark Tinka mtinka at africaonline.co.zw
Fri Jun 2 18:10:21 EAT 2006


On Friday 02 June 2006 14:15, Kondwani Masiye wrote:
> Hie list,

Hello.

> *** System received a Cache Parity Exception ***
> signal= 0x14, code= 0xa41b7803, context= 0x6258ea04
> PC = 0x603fc8dc, Cause = 0x4820, Status Reg = 0x34018002

Looking at this, it would appear your router experienced a crash 
due to a parity error.

Since you mention it's (randomly) continuous, and with the crash 
error message you get (*** System received a Cache Parity 
Exception ***), a search on the Cisco web site reveals the 
following possible cause and resolution (this applies more to a 
7500 VIP, but is the most similar circumstance to your 
situation):

ISSUE:

The following types of parity errors can occur:

    * Processor memory parity error exception (or cache parity
      error exception)
    * Packet memory read parity error
    * PCI bus parity error
    * Packet bus write parity error
    * CyASIC parity error
    * Reno read parity error
    * I/O controller parity error

Generally, if a parity error occurs only once, it is considered a 
soft parity error. 

RESOLUTION:

If a parity error occurs only once, monitor for another 
occurrence. At the first occurrence of a parity error, it is not 
possible to distinguish between a soft or hard parity error. 
Most parity error occurrences are soft parity errors and can 
usually be dismissed. If you have recently changed some hardware 
or have moved the box, try reseating said hardware. Frequent 
multiple parity error occurrences signify faulty hardware.

Soft parity errors are 10 to 100 times more frequent than hard 
parity errors. To greatly reduce the impact on your network, 
wait for a second parity error before replacing anything.

What IOS version are you running?

> Secondly, the box has two MICA cards with 120 modems each,
> making a total of 240 modems. It also has 8 E1 ports. With 6
> E1 lines connected to it, it does not seem to be able to
> handle more than 120 calls at a time. I have MRTG monitoring
> the number of calls vs number of free modems and it shows that
> the number of calls does not go over 120 (same thing for
> number of PPP sessions active).

I've heard of this problem before, but never quite experienced 
it. Then again, MICA modems have several issues.

Try using the latest, stable MICA portware code, which as of now 
is:

mica-modem-pw.2.9.5.0

Yeah, I know, it's old :-)... 2004.

> I have attached a configuration for the box and I am hoping
> that somebody can help to identify any misconfiguration the
> box might have.

I see you are using R2 signaling (non-PRI). I know it's an older 
protocol for delivery of E1 trunks. It does have some 
limitations when compared to PRI, but I haven't tested it with 
that many modems to know simulate your situation.

Let me dig deeper and let you know...

Mark.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 827 bytes
Desc: not available
Url : http://listserv2.cfi.co.ug/pipermail/afnog/attachments/20060602/d3f0c199/attachment.bin


More information about the afnog mailing list