Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

3368

December 2nd, 2016 11:00

What is a CA?

Hey guys. had an issue with our Compellant last week. Dell tech support pointed to stuck CAs causing excessive I/Os. What is a CA? What causes a CA?

Thanks,

vMojorod

1 Rookie

 • 

2 Posts

December 5th, 2016 06:00

Found the answer from Dell Tech Support.

  1. What is a CA?
    1. A “Config Access” is a ‘traffic cop’ that processes commands from other subsystems in the background.

 

  1. How does a CA become stuck?
    1. It is supposed to clear the processes after a period of time, when they don’t  they can be either ‘stuck’ or take longer than normal to clear.

                                          i.    In your case, they did become stuck or hanging around longer than usual,  but the controllers were able to clear them after a period of time.

 

  1. Is a stuck CA a normal occurrence?
    1. No it is not a normal occurrence, the system should clear them after period of time.

 

  1. If so, how do stuck CA’s get cleared in normal operation?
    1. If they become stuck,

                                          i.    Some instances the escalation team can clear them,

                                         ii.    In the instances the escalation team cannot clear them a controller reset may resolve it.

  1. In any case of a planned/manual controller reset, confirmation/system check from Copilot is strongly suggested.
  2. Also with production systems the last option we suggest would be a controller reset.

 

  1. Why didn’t the stuck CA’s get cleared in this instance?
    1. There were 2 stuck CA request – to change attributes on the eSX host that was being added at the time.

                                          i.    The reason behind this was due to an issue with the server not allowing to change the attribute

  1. Possibly an issue with the server or a FC fabric or FC switch zoning.
  2. The other CA requests  were getting processed normally
  3. Due to large amount of request happening in this instance, the system was not able to process all requests at normal speed.

                                          i.    Which lead to the performance degradation

 

  1. What actions can you identify that caused this? Is there a specific action in Enterprise Manager that causes CA’s to get sticky?
    1. The recent ESX host added was causing mapping issues with the SAN,

                                          i.    The reason it was having issues has not been determined.

 

  1. Is there a way to Monitor Stuck CA’s?
    1. Not at the customer’s end.

 

  1. How often to customers have issues involving stuck CA’s?
    1. Not often but it can occur, like getting a flat tire on to work.
No Events found!

Top