Jump to content

Recommended Posts

Posted

2 of our sites will require a SBC due to the number of basic Fanvil phones being used, however i'm not a fan of sticking it on our HyperV cluster as it's just another system to be aware of if we want to do maintenance etc. I want the phone system to be completely idependant of the school servers so it just runs it's self.

 

The information on 3CX for a SBC is a i7 8th gen processor. Is this spec really required if we are sticking linux on it?

 

One of the sites will have around wil have around 80 physical handsets, i'm not sure about the othersite as of yet.

 

I was hoping to get away with running a 2x HP 800 g2 mini PCs (i5 6500) in high availability at different locations acorss the site. Would this work or should we be looking st something more powerful?

Posted

I have been running our SBC on an HP 800 G2 for over 18 months with no issues. We have around 50 Fanvil handsets.

 

Hope that's helpful.

 

Ian

  • Like 1
Posted
1 hour ago, zoohead said:

I have been running our SBC on an HP 800 G2 for over 18 months with no issues. We have around 50 Fanvil handsets.

 

Hope that's helpful.

 

Ian

 

Ah that's good to know. The documentation says up to 100 - i7 gen 8 - https://www.3cx.com/docs/recommended-hardware-specifications-for-3cx/ - is this meaning up to 100 concurrent sessions / active calls? There is no way that we as a school will have 100 active calls. Across the trust we are only looking a 64 concurrent license. We are looking to us the Fanvil V61G for most desks with admin having the V63 (5x) and 1x X210 for reception.

 

44 minutes ago, MatthewL said:

Are Raspberry PI's no good anymore, they were the go to for SBC's at one point, I have seen many sites using them in the wild.

 

I've been told that Pi's will only support up to 10 basic phones which wouldn't be ideal for us.

 

Posted (edited)
23 minutes ago, timbo343 said:

I've been told that Pi's will only support up to 10 basic phones which wouldn't be ideal for us.

 

The HA SBC PI pair was for a site with 2 Fanvil X7, 7 Fanvil X6U, 100 Fanvil X3U - 109 phones total. It was a reasonable sized secondary, but you maybe using more features/more handsets that requires more power on the SBC?

Edited by ThomL
  • Like 1
Posted
13 minutes ago, ThomL said:

 

The HA SBC PI pair was for a site with 2 Fanvil X7, 7 Fanvil X6U, 100 Fanvil X3U - 109 phones total. It was a reasonable sized secondary, but you maybe using more features/more handsets that requires more power on the SBC?

 

Ta, that sounds a little bigger than one of our sites 👍

  • Like 1
Posted (edited)

In my old Trust we used to run the schools SBCs on mid spec (i5/8gb/256) HP laptops that we had spare.

 

In my new Trust, I moved them over to a new VoIP system and purchased 2 x SBC capable Yealink phones to handle the VoIP traffic. This allowed me to avoid having something setup locally. Works really well.

Edited by HyperTech
  • Like 1
Posted

Been using a pair of Pi SBCs at each of the sites I manage for the past few years, ranging from 10 - 70 phones (Grandstream GRP26XX series), with no issues at all.

  • Like 1
Posted

Honestly, I would still question why you'd not run it on your virtualised infrastructure - you need to maintain the SBC system anyway, and putting it on its own physical box just reduces the redundancy you have in place (that box fails, you've lost your phones, vs one of your virtual hosts failing, the VM moves to another host and you've lost nothing). We run ours as a VM and it uses very little in terms of resources, and we run about 100 extensions. The VM is currently using 1.2GB RAM, and barely any CPU at all (in Vmware terms, 359Mhz of CPU).

  • Like 2
Posted
1 hour ago, localzuk said:

Honestly, I would still question why you'd not run it on your virtualised infrastructure - you need to maintain the SBC system anyway, and putting it on its own physical box just reduces the redundancy you have in place (that box fails, you've lost your phones, vs one of your virtual hosts failing, the VM moves to another host and you've lost nothing). We run ours as a VM and it uses very little in terms of resources, and we run about 100 extensions. The VM is currently using 1.2GB RAM, and barely any CPU at all (in Vmware terms, 359Mhz of CPU).

 

It only moves if you have your Hosts set up with Failover / in a cluster 😒

Posted
27 minutes ago, timbo343 said:

 

It only moves if you have your Hosts set up with Failover / in a cluster 😒

 

We have ours replicating to another host so we can just switch it on if needed. Not much config required to set that up if you are using veeam

  • Like 1
Posted

We have one per school as a VM (2 cores, 4GB RAM. 40GB disk).

 

Building a new one and adding it to 3CX takes about 5 mins.

 

Looking at the 3CX HA page, I suspect you could make that work across two VMs on different hosts with a bit of faffing unless the setup script explicitly checks the hardware isn't virtual.

  • Like 1
Posted
20 minutes ago, Olliedawg said:

Slightly off topic question - how many of you run 3CX on windows vs on linux?

We’re running on Linux 

Posted

Our phone servers are Lenovo M720 tiny with i5 8th gen and 16 gb ram, because that's what we had lying around at the time. They aren't Entra joined and windows updates are basically manual at the moment.  Whilst we aren't a 3cx house any more,they were, and they were fine. They sit at about 20-30% cpu and about 60% ram running Windows 10 LTSC. We've liased with our telephony provider about options to move away from Windows 10 and they've advised that our phone server will be available as a Docker container shortly so we'll be swapping to Linux - probably boot2docker -  in the summer.

Posted

Just a thought, you can get VOIP handsets I think they call them router phones which means you don't need a SBC, I stand corrected if not but worth checking out.

Posted

This seems a bit backwards to me. Surely you have to "be aware of it" more on a physical machine? Hardware failures are much more likely to happen on a consumer / business desktop.

 

Your virtual infrastructure presumably has better levels of redundancy and resilience - dual PSUs, UPS, RAID / storage spaces direct, and hopefully some form of HA / clustering. For a service as critical as a phone system, it should be on your most resilient platform.

 

We run ours on our Hyper-V cluster so it auto migrates in the event of a host failure and when we apply patch Tuesday updates.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...