When we run "sh proc cpu" command, we see these outputs. what does ios-base process mean ?
ver : s72033-ipservicesk9_wan-vz.122-33.SXH7.bin
What we supposed to do to reduce cpu usage ?
Looking at your 'show proc cpu detailed 16407' it looks like high CPU is caused due to interrupts:
Backbone1#sh processes cpu detailed 16407
CPU utilization for five seconds: 57%; one minute: 57%; five minutes: 56%
PID/TID 5Sec 1Min 5Min Process Prio STATE CPU
16407 51.2% 49.4% 49.1% ios-base 10d06h
Process sbin/ios-base, type IOS, PID = 16407
CPU utilization for five seconds: 7%/44%; one minute: 6%; five minutes: 6%
Task Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Prio Task Name
So modular IOS displays this a bit different than traditional IOS...but when you look at the 5 sec CPU utiliztion and see X/Y. Y in modular IOS is the % of CPU caused by interrupts.
We have a guide that helps with troubleshooting high CPU:
The main part you will want to look at his here:
It should help you see what traffic is getting punted to the CPU. Once we know the traffic, we can see where did it come from, where is it going, and start to take a look at why its punted.
The best tool here is to span the CPU. This will give you an idea of what the traffic is, and that should give some clues to this.
see this thread
the best thing you can do is to move to standard IOS as explained by Edison and Reza.
Our forums have had several instances of cpu problems with modular IOS
I recommend to use standard IOS to my customer: if a new C6500 box arrives with modular IOS we install standard IOS on it.
Hope to help
first of all, thanks both of you for replies.
is it a good choice to upgrade SXI3 ?
what version do you suggest ?
with SXI, currently the latest is SXI4a. I have not test it 4 or 4a, but SXI4 came out a couple months ago. We found out that part of MPLS was not integrated into it and Cisco integrated the MPLS features and came up with SXI4a. So, if you have a lab, I would test 4a to make sure all the features you need are in there.
Also, as Giuseppe (Hi Giuseppe) already noted, go with non-modular.
We found out that part of MPLS was not integrated into it and Cisco integrated the MPLS features and came up with SXI4a.
Good info Reza (+5). Can you specify what feature(s) were added to the SXI4a code?
We wanted to run the SXI4a code but someone from Cisco has recommended our design architects to stick with SXI3 (much to our disappointment) because no large accounts are using it.
Thanks for the rating
The major feature that was added to SXI4 and 4a is that in a VSS environment now you can have 2 Sups per 6500. Also, there is a bug in SXI3 in regards to SSH key corruption when using eFSU to upgrade VSS pair. Basically when you use eFSU to upgrade a VSS pair the, SSH key gets corrupted and you lose connectivity to the device. Then you have to login via telnet and change the FQDN that was assign to the switch. Now, if you don't use eFSU everything works fine.
Thanks. What about the MPLS feature that was integrated to the new SXI4/SXI4a? Would you happen to have any information?
Sorry to get back to you so late
Here are some of the specific caveats that ware resolved in SXI4.
Regarding your question on MPLS features, there ware no additional MPLS features added in SXI4 The issue was that some of the MPLS features from SXI3 to 4 ware missed, so they added those missing features to SXI4a
* CSCsh51293-Resolved in 12.2(33)SXI4
The Secure Shell server (SSH) implementation in Cisco IOS contains multiple vulnerabilities that allow unauthenticated users the ability to generate a spurious memory access error or, in certain cases, reload the device.
* CSCsg65318-Resolved in 12.2(33)SXI4
Symptoms: Malformed SSH version 2 packets may cause a memory leak.
Conditions: This symptom is observed on a Cisco platform configured for SSH version 2 after it has received malformed SSHv2 packets. The impact of this flaw is that the affected platform may operate in a degraded condition. Under rare circumstances it may reload to recover itself.