High CPU utilization

[k_sinjish]

Hi ,

We  are getting this below error message at a  7200 router with ios   “7200-jk9o3s-mz.124-25b.bin

Error Messgae: Apr  9 09:51:02: %IP_VFR-4-FRAG_TABLE_OVERFLOW: GigabitEthernet0/0.129: the fragment table has reached its maximum threshold 16

We have identified the issue is because   of the datagrams being reassembled at any one time has reached it maximum limit. And the  recommended solution is to  set the maximum number  of the datagrams by entering  #   ip virtual-reassembly max-reassemblies <number>

  Because of  some administrative issue we could not go with this solution . And It would be  great if  you clarify  me the below  points.

1) If we are leaving the same issue as such (logs ) , what will be its impact in the device  ?
2) Recently  we   are getting  the high cpu utilization in this  router  very much  frequently. So please let me know the whether the above error  will cause for the high cpu utilization .

Thanks & Regards,

Sinjish.k

[Laurent Aubert]

HI,

Yes it's usually related as reassembly is only done in software (no hw support).

This feature is useful if CBAC is configured so it can inspect fragmented packets by reassembly them first and for NAT so it can have access to the payload (NAT ALG)

If none of those are configured, you can disable Virtual Fragmentation Reassembly on the interface. The other solution is to find out why you are receiving fragmented packets (MTU issue somewhere) and try to fix it.

http://www.cisco.com/en/US/partner/docs/ios/sec_data_plane/configuration/guide/sec_virt_frag_reassm_ps6350_TSD_Products_Configuration_Guide_Chapter.html

HTH

Laurent.