[c-nsp] 7201 cpu (revisited)

Tassos Chatzithomaoglou achatz at forthnet.gr
Sat Mar 22 08:44:31 EDT 2008


They were about 1000 sessions, having max 2 x 200/90 Mbps traffic passing through the box.

--
Tassos


Ben Steele wrote on 21/3/2008 2:33 πμ:
> How many PPPoE sessions did you have terminated and approx what traffic 
> flow in those graphs?
> 
> 
> On 21/03/2008, at 5:30 AM, Tassos Chatzithomaoglou wrote:
> 
>> We did some testing on a NPE-G2 for a week and this was the difference 
>> from NPE-G1:
>>
>> http://img84.imageshack.us/img84/905/g1vsg2px4.gif
>>
>> PPPoE termination, Qos/ACL/netflow per user, simple mcast, plus some 
>> basic routing stuff.
>> We used exactly the same setup (latest SB IOS) and same customers on 
>> both engines.
>>
>> G1 was usually showing drops after 70%-75%, G2 hasn't shown any yet.
>> If G2 can do anything above 75%, serving x1.5 customers (or x1.5 
>> traffic) of that of G1, without showing any drops, then we'll be
>> happy ;)
>>
>> -- 
>> Tassos
>>
>> Rodney Dunn wrote on 17/3/2008 8:54 μμ:
>>> Don't worry about it.
>>>
>>> Push more load and the CPU will go up but your overall no drop rate
>>> and performance is much more with that newer processor.
>>>
>>> The only way you can prove it is either in a lab with testing
>>> gear or watch from drops with the 'sh int'.
>>> You will see overruns or ignores once the capacity starts to come
>>> in to question.
>>>
>>> Not a good answer but it's the only one there for this scenario
>>> today.
>>>
>>> Rodney
>>>
>>> On Mon, Mar 17, 2008 at 11:52:03AM -0700, Mark Kent wrote:
>>>>>> Most of your CPU usage comes from interrupts.
>>>>>> sh int st
>>>>>> Lets see what your interfaces are doing.
>>>> GigabitEthernet0/0
>>>>         Switching path    Pkts In   Chars In   Pkts Out  Chars Out
>>>>              Processor      58937    6909233      53463    7238990
>>>>            Route cache 3561281996  824442297 1922951596  564241376
>>>>                  Total 3561340933  831351530 1923005059  571480366
>>>> GigabitEthernet0/1
>>>>         Switching path    Pkts In   Chars In   Pkts Out  Chars Out
>>>>              Processor      22461    2697899      27968    7456044
>>>>            Route cache  749956906 3834247096 1271047162 1998558616
>>>>                  Total  749979367 3836944995 1271075130 2006014660
>>>> GigabitEthernet0/2
>>>>         Switching path    Pkts In   Chars In   Pkts Out  Chars Out
>>>>              Processor          0          0          0          0
>>>>            Route cache          0          0          0          0
>>>>                  Total          0          0          0          0
>>>> GigabitEthernet0/3
>>>>         Switching path    Pkts In   Chars In   Pkts Out  Chars Out
>>>>              Processor     767250   64149215     783904   55907587
>>>>            Route cache 2991435297 3393180667 4107796986  931040113
>>>>                  Total 2992202547 3457329882 4108580890  986947700
>>>>
>>>>
>>>> Thanks,
>>>> -mark
>>>> _______________________________________________
>>>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>>>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>>>> archive at http://puck.nether.net/pipermail/cisco-nsp/
>>> _______________________________________________
>>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>>> archive at http://puck.nether.net/pipermail/cisco-nsp/
>>>
>> _______________________________________________
>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/
>>
> 
> 


More information about the cisco-nsp mailing list