Dynamic Queue Slots?

[PhoneGuyNJ]

Can anyone tell me what this means in our S8700? In our older R systems, you had a specified amount of queue slots, and had to put a number in on hunts to allot how many may be used by a split. Now I am able to put "unlimited" and the sys cap screen shows "Dynamic Queue Slots" as opposed to "Queue Slots Per System".

I need to clarify this for a client.

 

[yurichan]

Unlimited hunt queue slots imply that there is basically no limit to the number of calls you can queue to the skill. This a misnomer because there is a system limit. I believe it may be 12,000? You can see the system limit in the "display capacity" form. Look for something with "dynamic hunt queue slots"?

So let's say you have just 1 skill that is specified as "unlimited". You can have as much as 12,000 calls queued into that skill before you fail to queue calls.

Typically you use the "goto..if calls-queued(?)" conditional to check an upper bound of calls queued. Then route the call accordingly.

If you had 2 skills, both specified as "unlimited", then the 2 skills would share the 12,000 between them. If skill 1 happened to have 11,999 queued calls already, then there is only 1 available queue slot to share between either skill 1 or 2. Which ever skill the next call queues to will take that available queue slot. There is not pre-allocation. It is given as a first come first serve basis. After all 12,000 calls are sitting in queue(s), subsequent calls will not be able to queue. But if an agent happens to answer a call from the queue (any skill), then this frees up a queue slot for another call to queue to for any skill.


Your vectors that queue the calls should probably have a "queue-to skill" in a wait loop...

01 queue-to skill...
02 wait 30 secs hearing music
03 goto step 1 (don't goto step 2)...
..

Then if the call failed to queue the first time through the loop, it may be queued in a subsequent queue-to. As mentioned above, you may also want to check for an upper bound, because if you have over, say 100 calls in queue, you're probably not servicing the calls correctly and may need to requeue the calls to another skill with better service levels?


So there is a limit for each skill, you just don't know what it will be. Probably Avaya should have used "unknown" instead of "unlimited". Best case 12,000. Worst case 0. Chances are you won't hit the 12,000 system limit case. There are 8,000 incoming trunks and if all 8000 incoming trunks are queued to a single skill(s) you absolutely won't hit that limit. But if you queue to mulitple skill in a single vector..

01 queue-to skill 1
02 queue-to skill 2
03 wait ..
04 goto step 1 ..

Then a single call will take up 2 queue slots! In this case, if all incoming trunk calls queue to 2 skills, then 6,000 calls could sit in 2 queues each and will hit the 12,000 queue slot limit! The 6001st incoming call will not have an available queue slot to queue to any skill. Again, if you have 6,000 calls sitting in queue, then you got problems. Either you're agents aren't pulling calls off the queue to free up queue slots, or you may be queuing calls to skills that don't have staffed agents, or?

Hopefully this makes sense. A bit long? Sorry.

-y