I have a customer who has a problem where intermittently a call waiting in a queue will not ring an available agent. Their configuration is as follows. They have three queues where the same agents are members in all three queues with penalties of 0. Each queue is set for ring all ring strategy. Queue 1 has it’s queue weight set to 9. Queue 2 has it’s queue weight set to 7 and Queue 3 has it’s queue weight set to 0. The queues do not have any wrap up time set. The problem seems to only happen with calls that go in to queue 3. Intermittently it will show calls waiting in queue 3 and one or more extensions will show not in use but the phones do not ring. It seems to happen after the extensions finished up calls from one of the other queues but I can’t confirm that.
Asterisk only allocates calls to queues when the call polls for an agent (once a second when the call isn’t playing a comfort message). That means you can have a situation where there is a call queued for a higher priority queue when a call in the lower priority queue next polls. If the lower priority queue was allowed to grab an agent, you could potentially have a situation where the high priority call misses many turns, because agents are becoming free just before low priority calls poll. To get round this, Asterisk does a check, when allocating an agent to a low priority call, to see it the agent is also in a higher priority queue, with a call pending. The agent won’t be marked as busy, if you look at the agent states, but will be treated as though they were busy.
I think that is still the case, although I haven’t checked.
I wonder if that is what you are describing.
Thanks for the quick reply. In our case it’s the calls in the lower priority queue that are not ringing the available agents so I don’t think that matches what you are describing.
Sounds like it is matching. The low priority calls are discovering that their agent will be needed to cover a high priority call.
Maybe I am missing something, but this happens even when there are no other calls pending in the higher priority queues.
Ok. That is new input.
Sorry, that wasn’t clear. It acts like there is a wrap up time set on the other higher priority queues because it seems to happen after the agents were on other previous calls in the higher priority queues. Their queue status will “not in use” in all the lower priority queue but the phones don’t ring and then after a minute or two they will start ringing again. It’s very intermittent.
My suggestion is to start from queue_log and document a case when this happens, and then go back and see what was the status of agents at the time.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.