[c-nsp] QoS / LLQ crashes NPE-G1 with 12.2S

Brian Feeny signal at shreve.net
Tue Mar 1 21:19:25 EST 2005


I have done some pretty extensive testing, on both 12.2.18S7 and 
12.2.25S, and I am pretty sure enabling
QoS on a few interfaces on one of our 7206 is causing that box to 
reload with:

dal-lone-gw1 uptime is 3 hours, 20 minutes
System returned to ROM by error - a Software forced crash, PC 
0x60763C94 at 23:00:45 UTC Tue Mar 1 2005
System restarted at 23:03:59 UTC Tue Mar 1 2005
System image file is "slot2:c7200-p-mz.122-18.S7.bin"


I do have a ticket opened, and now have 5 crashlogs on this.  The 
policy-maps I have are quite simple:

class-map match-any Mark-VoIP-RTCP
   match ip rtp 16384 16383
   match access-group name IAX-Traffic
class-map match-all Queue-VoIP-Control
   match ip dscp af31
class-map match-all Mark-VoIP-Control
   match access-group name SIP-Control
class-map match-all Queue-VoIP-RTCP
   match ip dscp ef
!
!
policy-map Mark-DSCP
   class Mark-VoIP-RTCP
    set ip dscp ef
   class Mark-VoIP-Control
    set ip dscp af31
   class class-default
    set ip dscp default
policy-map QoS-Policy
   class Queue-VoIP-RTCP
     priority 10000
   class Queue-VoIP-Control
    bandwidth percent 5
   class class-default
    fair-queue


Basically I use "Mark-DSCP" for inbound stuff thats untrusted, and 
"QoS-Policy" is my outbound policy.
The router can last 24-72 hours tops before it reloads.

I have many other routers running those some maps, same IOS, but are 
NPE-300's, and not necessarily the same
PA's.

Has anyone else seen this or is this something new?

Brian

---------------------------------------------
Brian Feeny, CCIE #8036, CISSP
Network Engineer
ShreveNet Inc.



More information about the cisco-nsp mailing list