Meridian 1 relation between NCR and CSQI/CSQO 1

[DanyV]

Hello everyone,

We're having some communication problem between the MLINK and the M1.

I'm looking the configuration and I have to modify the value of NCR and CSQI/CSQO.

I know that NCR means Number of call registers and the CSQI and CSQO corresponds to the input and outbut buffers.

But what is the relation between NCR and CSQI/CSQO?

Thank you for your help!!!

 

[diveley]

The CSQI AND CSCO are the call registers for the incoming and outgoing link queues. The number of registers is based on a percentage of call registers configured for the Meridian 1/SL-1 call processing. The recommended setting is 20, or 25% of the system call registers allocated, whatever is larger. Nortel has recommeded call register sizes depending on switch type and software release. To change the call register size go to LD 17 under parm prompt ncrmake change after a change is made here the switch has to be initialized for the to take effect. The CSQI and CSCO prompts can then be increased to 25% of the new call register size.

 

[rhack5565]

What type of problem are you having? The type of issues I have seen with the CSQI/CSQO being too low had to do with aquiring resources on the M1. For instance Symposium has to "aquire" the ACD sets under its control. With the CSQI/CSQO set low, not all sets would acquire. Multiple attempts would have to be done to do so. I know back in Rls 24, there was a patch on the M1 that in essence created a multiplier for the CSQI/CSQO buffers that allowed a far greater number of buffers to be reserved.

 

[DanyV]

Hello,

I'm having these errors coming from the PBX:
CSA 100 O 8 34 48 AML 9 S 4, AML 10 S 812, AML 11 S 0
CSA 101 O 8 35 04 AML 9 S 0, AML 10 S 1020, AML 11 S 0

Everything comes back to normal:
CSA 102 O 8 37 42 AML 9 S 0, AML 10 S 612, AML 11 S 0

On the Meridian Link side, I have this error:
04/03/02 08:31:24 hlinklan: msg send retry limit exceeded

I read the Appendix E of the Modular Option EC Installation and Maintenance Guide. But I did not find any explanation about what is the relation between NCR and CSQI/CSQO.

I know the definition of each but if I change a value, wich impact does it have on the other one?

Can you (diveley or rhack5565 or anyone else) give me more details?

I don't want to change any value before knowing what exactly I'm doing.

 

[rhack5565]

As far as the relationship goes NCR's are your total "pool" of resources. As diveley said, the rule of thumb is for your CSQI & CSQO to not exceed 255 of your call registers. There is another register setting "MGCR" which is used for passing ACD related messages over the High Speed Link to Meridian Max. These are independent of the NCR's.
If you are using both Mermail and ACD sets that Are AST sets (controlled by MLINK) you will need the following:
Go into LD 81 and do a CNT for FEAT ACD. Use this number as your minimum CSQI & CSQO numbers. If the numbr is more than 25% of your NCR's, increase your NCR's accordingly, then increase your CSQI/CSQO. I would pad at least 10-15% to allow for increases of agent sets. Like dliveley said, you will have to initialize the switch after making the changes. Its possible you are seeing the Hlinklan send issue as a byproduct of the link not being able to send a message to the PBX.


What type of system do you have? Option 61C, Option 81C etc..
What is the Generic/Release & Issue of software?
What are your current NCR and CSQI/CSQO settings?

I take it you are seeing these errors during high traffic times? What are you using MLINK for? What is the third party application connecting through the link?

Maybe we can help with some suggested settings.