Cisco 6509 CPU Usage

Posted on 2011-04-29
Medium Priority
Last Modified: 2012-08-14
We have allot of users that are complaining about slow performance.  Our   two 6509es that have memory around 80% and CPU is averaging about 65%.  That seems high to me.  any suggestions on where to look?
Question by:japplewhaite
  • 2
LVL 10

Accepted Solution

koudry earned 500 total points
ID: 35493035

The Cisco 6500 platforms, are high end devices designed for core network and equipped to hand huge amount of traffic without being stressed. However, I have seen a scenario when a Cisco 7606-S router crashed before my eyes.  Both 6500 and 7600 are quite similar.

In my case, the 7600 router crashed because of high traffic generated using SmartBits test engine and packet fragmentation.  Fragmentation occurs when the packet received by the router has a higher MTU than the MTU configured on the outgoing interface.  

A typical scenario is, you are receiving a gigabit traffic on a gigabit port, the input port has an MTU of 2000 bytes and the output interface has an MTU of 1500 (default).  The CPU fragments each packet into 2 to be able to send it via the output interface. This fragmentation makes the CPU works harder, so the CPU usage goes up, in my case, it went to 100% and the RSP crashed.

To avoid fragmentation, check and make sure that MTU is the same from end to end.

The other feature that can drive your CPU usage up even on high end routers, is VPN.  A platform like 6500 is equipped with a dedicated VPN hardware.  The 6509-E has 9 slots one of which can be a VPN concentrator.  However, as the number of VPN tunnels goes up, CPU usage goes up to and performance goes down.

This is just a thought based on my experience.

LVL 29

Assisted Solution

by:Jan Springer
Jan Springer earned 1000 total points
ID: 35493192
sh proc cpu sorted


sh proc mem sorted

when the loads are high
LVL 12

Assisted Solution

Fidelius earned 500 total points
ID: 35494302

Also if you are monitoring your 6500 via SNMP and have a big routing table (especialy if it is a BGP peer) it can use quite a lot of CPU. In that case you can limit SNMP readout to only necessary OIDs, or filter out routing OIDs from polling.

LVL 29

Assisted Solution

by:Jan Springer
Jan Springer earned 1000 total points
ID: 35494384
Yes, with the SUP processors, BGP can spike the CPU.  The RSPs are much better at handling that.

Featured Post

Free Tool: Port Scanner

Check which ports are open to the outside world. Helps make sure that your firewall rules are working as intended.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

I see many questions here on Experts Exchange regarding switch port configurations and trunks. This article is meant for beginners in the subject to help to get basic knowledge about Virtual Local Area Network (VLAN (http://en.wikipedia.org/wiki/Vir…
In this article, the configuration steps in Zabbix to monitor devices via SNMP will be discussed with some real examples on Cisco Router/Switch, Catalyst Switch, NAS Synology device.
We’ve all felt that sense of false security before—locking down external access to a database or component and feeling like we’ve done all we need to do to secure company data. But that feeling is fleeting. Attacks these days can happen in many w…
Please read the paragraph below before following the instructions in the video — there are important caveats in the paragraph that I did not mention in the video. If your PaperPort 12 or PaperPort 14 is failing to start, or crashing, or hanging, …
Suggested Courses

850 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question