[c-nsp] 6500 Sup720 high CPU load - RP LES Fragmentation unsupported

[Sukumar Subburayan]

comments inline..

sukumar




On Mon, 4 Dec 2006, Bernhard Schmidt wrote:

> Sukumar Subburayan wrote:
>
> Hi,
>
>> The new CEF Fast Forwarding feature supports fragementation of IP packets 
>> in CEF-path.
>> 
>> The 'RP  LES Fragmentation unsupported' is the number of packets punted by 
>> CEF to the process level as it was not able to fragment the packets for 
>> whatever reason. I see there are some internally found/fixed bugs which are 
>> not in 12.2(33)SRA release, and will be available in an upcoming release.
>
> Do you have any BugIDs for that? I've been on site today and the traffic is

There are a bunch of bugs, which have been duped to an internally 
found/fixed bug. Since, they are internally found bugs you cannot view 
them. That is why I suggested opening a TAC case and have someone verify 
if you are running into the bug, before moving it to 'customer-use'.

Anyway, the bug fix is not going to fix anything major, except handling 
the packets needing fragmentation in CEF fast-path.

You still may want to fix the application.

> generated by a (supposedly) buggy NetApp-Filer which creates NFS-over-TCP 
> packets of 1512 bytes (IP) without df-bit set. I guess the box tried pMTUd as 
> a start and got some wrong idea about its own overhead (forgot the 12 bytes 
> of Options it sent in the TCP header, doh!)

>
>> Once thing is not clear is, how the router is receiving > 1500 byte packets 
>> when you say that on the other end there is only one IP interface, which 
>> has mtu at 1500 bytes.
>
> The physical ports (6704-10GE and 6748-GE-TX) are configured to MTU 9216 as 
> well, but the SVI is configured to 1500 Bytes. I guess the size is not 
> checked on incoming packets on SVI.
>

OK, this makes sense. When MTU is configured to 9216 on the ingress, we 
will check to see if the packet is less than that on the input interface 
(the 6748-GE-TX linecard) and when we try to L3-switch the packet, we see 
the SVI interface is 1500 bytes. We probably are punting the packet to the 
RP for fragmentation.

Can you check to see if you configure 'mtu 9216' on your SVI interface, 
that the CPU comes down?


  >> I would suggest opening a TAC case and investigating this 
further. >
> Working with our distributor is always a challenge in this regard, but I'll 
> have a look at that.
>
>> Something things to consider when working with TAC is :
>> 
>> What type of linecard is on the other side on Vlan71?
>
> The traffic I've seen was incoming on WS-X6748-GE-TX. Interestingly there are 
> no giants listed on the interface in "sh int", but the giant counter on the 
> other router (where ingress is 10GE routed) is increasing. Different IOS 
> though.

There are some differences in the way 'giant's are reported on the 
linecards and particularly if the ports are configured as trunk-port vs 
switchport.

For ports configured as trunk ports, we will not report giant packets for 
packets upto the size of 1548 bytes.

>
> Thanks,
> Bernhard
>