[c-nsp] 7206VXR NPE-G1 Upgrade from 12.4 to 15.0 High CPU

[Arie Vayner (avayner)]

You are referring to NPE-G2 (or 7201)

Arie

-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net
[mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Chuck Church
Sent: Thursday, August 25, 2011 04:58
To: 'Chris Gotstein'; 'Joe Maimon'
Cc: 'cisco-nsp'
Subject: Re: [c-nsp] 7206VXR NPE-G1 Upgrade from 12.4 to 15.0 High CPU

Wasn't the G1 the one that Rodney Dunn mentioned would perform well at
high CPU load, and the CPU usage didn't increase linearly with the PPS,
or was that the G2?  Personally, I had a G2 running 15.0(1)M5 doing
complicated SLA/policy routing, and 50 megabit didn't move it past 3 or
4%, 2 peers with maybe 10,000 prefixes (full DISA table).

Chuck

-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net
[mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Chris Gotstein
Sent: Wednesday, August 24, 2011 7:57 PM
To: Joe Maimon
Cc: cisco-nsp
Subject: Re: [c-nsp] 7206VXR NPE-G1 Upgrade from 12.4 to 15.0 High CPU

We had definite issues when the CPU starting getting above 80%.  At 90%
it was almost unusable.

On 8/24/2011 5:02 PM, Joe Maimon wrote:
> I have seen this in other instances as well. A 2811 never dips below 
> 10% (cpu is graphed) even off-peak with nothing going on. The 1700 it 
> replaced properly flat-lined to 0% at the same times.
>
> It remains to be seen whether this is an aesthetic issue or an actual 
> problem.
>
> I am looking at a G1 running 15.0 pushing a few hundred mbps peaking 
> at near 90% (cpu graph) without any user reported issues reaching me. 
> Some of that traffic is IOS FW, some is ATM, most is inter-vlan. Full
tables.
>
> Joe
>
> Chris Gotstein wrote:
>> Decided to move from IOS 12.4 to 15.0 (15.0(1)M7) on a 7206VXR with 
>> NPE-G1. Since doing the upgrade, i've seen increased CPU usage in the

>> IP Input process. The configuration remained the same after the
upgrade.
>> Verified CEF on all interfaces. Followed troubleshooting docs on 
>> Cisco's site, as well as searching the archives of nsp. Not able to 
>> find any info to help troubleshoot this issue. This is an edge router

>> connecting to our 2 upstream providers, taking full BGP routes. As of

>> 4am (CST), traffic is pretty much dead, less than 20Mbs going 
>> through, and CPU usage is at 35%. Normally it's around 10% for this 
>> time of night. Any ideas of what could be going on? I haven't 
>> downgraded the IOS just yet, hoping to see if i might have missed 
>> something easy. Thanks,
>>

--
---- ---- ---- ----
Chris Gotstein, Network Engineer, U.P. Logon/Computer Connection U.P.
http://uplogon.com | +1 906 774 4847 | chris at uplogon.com
_______________________________________________
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/