Coming from the FreePBX community. I’ve been looking over the software for the last few months, and have been really impressed with the rate of innovation and the apparent focus on call centers especially. The phones (both windows and VitXi RTC) both seems to have contact center call controls (wow!), and the commercial suite of tools looks really well thought out.
We have brought to bare various queue solutions, albeit with some difficulty, and it looks like, maybe, with VitalPBX, it does not have to be quite so complicated. So at the risk of asking unanswerable question, I want to know about queues capacity. How much simultaneous volume can a queue handle? How many agents can a queue handle? I know this is largely hardware dependent, but I’ve had 36 core, 64 Gb Ram servers choke on a surprisingly small number of agents and queue calls with the FreePBX queue logic.
So maybe a more answerable question is, what’s the largest queue in terms or call volume and/or agents that you’ve deployed, that’s worked well?
As you have mentioned, this will be highly dependent on your hardware. With VitalPBX and many other Asterisk-based PBX systems, simply adding more CPU cores to the server will not improve performance.
There is no exact formula for this as it will depend on whether or not you are recording the calls, or if you are compressing the calls using a codec like G729. You can read more about this topic in this article we have made along with a script to measure the number of concurrent calls your PBX hardware can handle, What is the Concurrent Call Capacity of VitalPBX or Asterisk PBX? | VitalPBX - Advanced PBX System.
Based on some tests we have made, and hope to publish soon, the maximum server specifications would be 16 cores, 32GB of RAM, and at least 4TB of storage. Having a 32, 64, 128 core system will bear no difference. Given that Asterisk is a Multi-Thread application, not Multi-Core. Some benefits are shown from adding cores, but this is reflected as adding more threads and there’s still a limit within it.
In layman’s terms, we haven’t tested this directly on a queue per se, but a system as mentioned above will allow you to have up to 5,000 extensions at around 1,000-1,200 concurrent calls. And we have various customers with over 50 queues and dozens of agents that don’t report any issues that we know of.
I went into a bit of a tangent there, but hopefully, this gives enough context and is able to answer your question.
Exactly what I was hoping for, thanks for this. And only further impressed by this community that you even have a test script, very helpful. I am amazed at the progress this platform has made in such a short time. I will be setting up labs in the coming weeks and give this platform a go. Exciting.