show commands not showing - ACE module.

Unanswered Question
Jan 15th, 2009
User Badges:


After a few days of working fine the systems stopped answering SNMP and then the CLI began having problems with some commands. The easiest to see being "show running-config". At this point it shows nothing. It just waits and does nothing. I have to hit Ctrl-C to bring back the prompt.

The applications being balanced are working fine but the management is not.

The resources were allocated this way:

resource-class ADMIN

limit-resource all minimum 6.00 maximum equal-to-min

resource-class CLA

limit-resource all minimum 46.00 maximum unlimited

resource-class MAS

limit-resource all minimum 24.00 maximum equal-to-min

resource-class OL

limit-resource all minimum 24.00 maximum equal-to-min

The version running is A2(1.3) and the same thing happened with an earlier release.

It's a HA scenario and the problem appeared so far only in the CLA context which was active on different modules and did the same on both. That was just a test to discard hardware problems.

One of the messages we saw was:

ACE-CER/Admin# show resource usage summary

Error: Called API timed out

We are unable to show resources, running-config, etc. for the active context CLA.

Do you have a suggestion? Is it healthy to allocate “unlimited” as the upper limit? Could this be the problem?

Thanks a lot.


  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Gilles Dufour Thu, 01/15/2009 - 09:51
User Badges:
  • Cisco Employee,


this is usually a problem when your config gets really big.

Is that the case ?

You should definitely open a service request with the TAC because we are currently tracking those kind of issues.



ggalteroo Thu, 01/15/2009 - 10:43
User Badges:


If you mean a “get” for a probe, in this context I only have one and it's not that big. It goes like this:

request method get url

On the other hand, in a different context I have nine of these:

request method get url

3x, 6x and 16x mean that amount of letters.

Are these long enough?

To your knowledge, is this condition going to spread further and disrupt production traffic?

So once you have reached this condition whether caused by "gets" or not, would you recommend opening a TAC case?

Thanks a lot!


lalvenka Fri, 02/06/2009 - 09:07
User Badges:

Hi Guido,

The problem "Error: Called API timed out" could also be due to the logging level set to 5 or 7. If yes, then please reduce the logging level to 3 and see if that resolves the issue



ggalteroo Fri, 02/06/2009 - 09:36
User Badges:

Hello Lalitha

I was advised by a TAC engineer to do exactly that among other things. It could be a combination of resource allocation and too much logging. He told me to set a level 4 for logging and to rearrange the resourses differently among contexts. At this point we are waiting to see what happens.

Thank you!



This Discussion