Genesys CTI capacity problem

[cuneyt2919]

Hi all,
We have basic problems about CTI integrations. We use BC10 CN144 with 24 lim system and have Genesys Framework 6.5 version..WE have 450 desks monitored and 600 IVR hunt group members monitored.

Now, sometimes,on a lot of agent's desktop applications no cti data is retrieved. Or the status of agents reported wrongly to T server Statserver.... And by checking genesys tserver logs i can see some errors such as
generic system resource availibility
Outstanding request limit exceeded ...etc.


What i did is, last month i added new NIU cards,(NOw i have 24NIU cards)and changed the ethernet port 10MB-half duplex to 10Mb full duplex.

What i want to learn is,there are some parametres on Tserver side,Does anybody here tell us the true values, or are there any limitations on MD side for these parametres.
Or which parametres do you use for your system and of course for how many extensions do you use these parametres.

*****************************************

link max-request-rate (maximum number of requests per second that can be sent to the switch.. default value 60)

link maximum outstanding device request maximum number of requests that can be sent to the switch on behalf of a phone set before recieiving any confirmations or rejections.Any additional requests are queued. Default value 1)

link request delay (specifies the interval in miliseconds that tserver waits before attempting to send a feature request to the same device default value 200)

link maximum outstanding request ( maximum number of requests that can be sent to the switch before receiving any confirmations or errors default value 4)

link request timeout (Specifies the time interval in seconds, that genesys waits for request confirmation or error default value 20)

force snapshot interval (Defines a timeout in miliseconds after which Tserver will query the switch for the status of devices after completion of transfers in which the primary call was on ADN and consult call was on ODN default value 1000)

device snapshot interval Specifies the time in seconds that Tserver scans the list of devices for their states default value 60 , maximum 120 and we use 90 "0" (zero), turns the option off)

 

[jerommeke]

Hi cuneyt2919,

I've seen the errors and found out there was no problem with T-Server but with the application link.

I checked the Genesys ticket-history and found the same conclusion.

When I had this issue, in a much smaller envrionment using G6.1, we have removed "ip-security" from the applink to improve performance and solved the issue.

Also using the latest applink software and update CME parameter app-link-version.

May be an open door but is the app-link computer capable to handle all I/O and CPU requests.

If network traffic is suspected you could isolate network segments:
MD110 <-> Applink,
Applink <-> T-Server,
T-Server <-> Client

Keep us informed

regards,

\\Jerommeke

 

[cuneyt2919]

Hi jerromeke,
I think you're right, it's about network issue. I saw that our Tserver has 10Mbit half-duplex ethernet port
And also Application link had 10 Mbit half-duplex.
And we saw lots of data collisions.

Firstly, we changed Application Link ethernet port 10 Mbit full-duplex
and this week we will change T server port to 100Mbit...

Sometimes before t-server crashes, i saw that, the requests could not reach to Application Link.. And on t-server logs i saw "time out operation" error. Also when i increased the traffic rate i saw "network congestion error" on t-server logs.
Also during problem occurs, T server desktop client could not connect to CTI through the Genesys T server..
But i tried to connect to Application Link via NCSTAtool..
And could login to the system with no problem.
So i saw that T server produces a high network traffic.And 10Mbit half is not enough for it.
Especially there are critical parametres, such as device snapshot interval parameter...
When i changed it to 120 seconds lots of agents are seemed to be in wrong state according to T server , on other say T server can not realize that agent is logged off. And i saw lots of error such as (Agent Id is used on XXXX extension"
This is because,T server can not update the info for the agents if this parameter increased to 120 seconds...
When i decreased this parameter to 60 seconds, that time Application Link or DCTAP or DCTPP -not sure of it-,could not work properly... I saw CSTA link errors on MD110 side...(Not for all links but for some of CTI links.)
I think, this is because, on 60 second range Md could not response immediately...
So now i try this on 80 seconds.
For other parameters , i tried lots of changes, Actually i am not sure the problem is on md traffic or on Application Link or on Network communication failure

I will inform you after 100Mbit change for T server..

Regards