Question about UCCX callback script

[JMIII]

I have a callback script and have a question. When it gets to the place call to the callbackscript trigger it will go to successful even though someone in the callback csq has not gone ready yet. In reactive debug ill watch it keep bouncing back to the wait_for_agent_callback tag as that is listed for this menu step in the timeout and unsuccessful. When the agent goes ready it completes the call back and all works fine. My main concern is the limitation on steps. So if someone requests a callback and i only have that 2 second delay it will eventually reach the max 1000 steps right and caller will never get the automated callback correct? Is there a best practice to deal with?

To add an update i have done some further testing leaving a call waiting on a callback for an extended period of time and it never fails. So i am unsure if what i see when doing reactive debugging and what is actually occurring is different as i'd expect how it keeps stepping in the reactive debug to eventually lead to too many steps?

[Emma-Kate]

Did you solve this problem?

I am interested in what is the longest time you were able to keep calls in the call back queue before the script reached its maximum steps and abandoned the calls? 

[Anthony Holloway]

Keep in mind that if your place call step is in a script where the caller was originally waiting in queue, then all of those steps execute before will count against you. Some callers may chew through 500 steps before getting to the place call step, and others might go through only 50.

Also, keep in mind, that its also your other script (the one the Place Call Step calls) which you need to watch out for max steps too. Granted, it gets its own 1,000 steps, but it is a consideration. In fact, one trick you can do to ensure this never even happens, is to get rid of the queue loop in favor of a delay step (no music needed either) set to 12 hours (the system maximum).

Another thing I would recommend is to not have the Agent press a key, rather, in the called script (where you now have your 12 hour delay), use the Set Enterprise Info step to flag the call as answered by the Agent after its....well...answered by the Agent. And then in your calling script (the one with the Place Call step), you're looping a polling check with the Get Enterprise Data step, to see if the flag is set. This should give you an easier number to calculate. For example, if you poll every 10 seconds, and it takes 4 steps to poll: delay, get data, If and Goto, then you could theoretically hold the call in this pattern for 1,000max / 4steps * 10sec / 60sec = ~40min. And bumping it to a 15 second polling interval, nets you a ~60min hold time.