Nexus 7000 copy run start "config change is already in progress"

Answered Question
Jan 14th, 2011
User Badges:

Hello,


we have an N7k with redundant supervisors.

After a copy run start command was aborted (ctrl-c) because it took to mutch time, we can no longer save te configuration.


n7k# copy run startup-config
Configuration update aborted: another request for config change is already in progress
n7k# sh startup-config
configuration change in progress


We tried reloading the standby supervisor, but it won't boot anymore.


The switch is only live for 4 weeks and i hate to have to powercycle it already.


Does anyone have an idea how to recover from this?


Thanks,

Jacco

Correct Answer by habadr about 6 years 5 months ago

Hi Jacco,


What is the NX-OS version that you are running. I think you are hitting a bug (CSCtj44206).


From the bug toolkit


"N7K-reg : 5.1.1 S37: copy r s aborted due to timeout.


Symptom:

The utaker klm queue overflows. A syslog of this format will be seen in the supervisor's show logging output.


%KERN-2-SYSTEM_MSG: Utaker overflowed. Size -40/5242880 - kernel


Conditions:

One of the conditions when this can happen is when a large number of processes exit/crash.


Workaround(s):

Reloading the relevant supervisor is the workaround. If the issue is seen on the active sup, then perform a switchover. If seen on the standby sup, reload the standby module."


It is fixed in 5.1.2


Thanks


Hatim Badr

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 3.5 (2 ratings)
Loading.
Correct Answer
habadr Fri, 01/14/2011 - 06:30
User Badges:
  • Cisco Employee,

Hi Jacco,


What is the NX-OS version that you are running. I think you are hitting a bug (CSCtj44206).


From the bug toolkit


"N7K-reg : 5.1.1 S37: copy r s aborted due to timeout.


Symptom:

The utaker klm queue overflows. A syslog of this format will be seen in the supervisor's show logging output.


%KERN-2-SYSTEM_MSG: Utaker overflowed. Size -40/5242880 - kernel


Conditions:

One of the conditions when this can happen is when a large number of processes exit/crash.


Workaround(s):

Reloading the relevant supervisor is the workaround. If the issue is seen on the active sup, then perform a switchover. If seen on the standby sup, reload the standby module."


It is fixed in 5.1.2


Thanks


Hatim Badr

jbraat Sun, 01/16/2011 - 23:22
User Badges:

Thanks Hatim,


this looks like the problem we have. We are running 5.1(1a) and i see a "utaker" message in the logs.

However the sollution in the bug toolkit about reloading the supervisor is a problem. We tried reloading te standby and now it is stuck in powered-up state. So we can't switchover and restart the current active supervisor anymore. I guess we have to restart the chassis.


Again thanks for the input,

Jacco

tschaffran Tue, 01/25/2011 - 11:31
User Badges:

Has any other workaround been identified to kill the hung process?  I am facing the same issue, but I fear my

startup-config may be corrupt and when I power cycle the chassis, it will boot into an unconfigured state.

srinraja Thu, 08/02/2012 - 04:13
User Badges:

Hello All,


I have similar problem where difference of running config are not copied to startup.

Device having OS version with Nexus7000 5.0.3. Am i hitting the same bug got fixed in 5.1.2 version?


Sh run

======

version 5.0(3)

boot kickstart bootflash:/n7000-s1-kickstart.5.0.3.bin sup-1

boot kickstart bootflash:/n7000-s1-kickstart.5.0.3.bin sup-2


Thanks

Srini R

Actions

This Discussion

Related Content