12 thoughts on - PBX Selection

• I prefer Asterisk for my projects.

  Hire a team with knowledge about VOIP, without your prefer if you use Asterisk or whatever you want You will win a brand new full responsibility with VOIP. The learning process is long and hard. You will find a lot of problems like NAT, intrusions. Consider learn before you pain this.

  Yes.

  did you google about this?

  You need some skills but today is really simple.

• Hi Speed Boy.

  I agree with Emiliano Vazquez too.

  Additionally, you and your team must think others points before choose Asterisk:

  * Asterisk is build to work on Linux. So your team needs some skills like setting up a basic Linux server (Debian, CentOS, etc), donwload software from Internet, compile and install software manually.

  * Your team must know how to configure Linux networking. And solve NAT
  issue if apply. Basic network protocols like UDP, SIP and SDP/RDP are welcome.

  * If Asterisk needs interact with external world via VOIP provider, then you must know how to configure SIP or IAX2 trunks. If you have analog (like FXO) or digitals lines (like ISDN or similar), then you need ti know how to install and configure hardware on the Linux server like telephony cards
  (PCI-e or PCI) or configure VOIP gateways.

  * Security: How to install and configure a basic firewall (using iptables), o Fail2Ban. And best practices in Asterisk about this topics.

  Cheers

  El 17 abr. 2017 13:03, “Emiliano Vazquez”
  escribió:

• It may be that the developers mostly use Linux but Unix (i.e. BSD) works perfectly fine as well. I run it on NetBSD and it is rock solid.

  As for Asterisk vs. FreeSwitch, as a data point I started out with the latter but converted to Asterisk before going live. FS really was more Linux oriented and I had many problems getting it working on NetBSD. I
  also find the support here better.

  If you run NetBSD (and probably any other BSD) you can install from their package system. I compile from source using the package but you can also install a pre-compiled version just as easily.

  Cheers.

• Proponents of proprietary solutions always like to say “If an Open Source solution breaks, who can you call?” The answer is, “Any sufficiently-competent programmer — it may be broken, but we have all the pieces”. Whereas if you spend money on proprietary software and it breaks, then there is only *one*
  place you can call — and you’d better hope they are interested to fix your problem.

  On the other hand, if you could get full Source Code and Modification Rights
  (basically, “everything we could do with a GPL program except distribute copies”), a proprietary solution might not be so bad after all. But since the goal of most proprietary software vendors is to extract money from you and maintaining you in a state of perpetual helplessness is highly desirable in the course of this, do not expect to get such a deal in real life.

  Yes.

  Not sure about this. We’re still using the original chan_sip driver.

  It’s about as easy as compiling anything from Source Code. Harder than LAME
  MP3 encoder, but easier than the Linux kernel. If you altered `monop` from the BSDgames package to make the streets match your local edition of the game, you will have no problem whatsoever with building Asterisk.

  If you understand the process of what you are doing — basically, setting up an automated process that will examine your server hardware and software configuration (configure), choosing which parts of Asterisk you want to include (make menuselect), compiling the selected human-readable Source Code into binary code that the computer can understand natively (make) and then moving the compiled binary code and configuration files from the Source Code folder to where the computer is expecting for them to be (make install) then you should not have too many problems.

  It is always preferrable to compile your own Asterisk to fit your hardware and include just the bits you want, rather than rely on anyone else’s pre-compiled package.

  The latest one.

  You can certainly use Windows softphones to talk to Asterisk, but Asterisk itself requires a non-toy underlying operating system. Ubuntu and CentOS are the best-supported Linux distributions. Asterisk has also been seen working, to greater or lesser extents, on Solaris and the BSDs. But Linux was the original development environment (although one of the two original projects that ended up merging and becoming Asterisk, many years ago, was originally developed on FreeBSD), and is what most Asterisk telephonistas know.

  Any hardware which is capable of running Windows can, of course, run Linux;
  and usually better.

• Feel free to take a look at https://github.com/lardconcepts/asterisk-digitalocean-voipfone-config/blob/master/Asterisk-14-on-Ubuntu.md

  Ignore the bit about Voipfone and just skip to the “Install Asterisk” bit.

  I’ve used this same script with Asterisk 12,13 and 14 on Ubuntu 15,16
  and 17 so this should work!

  Let me know how you get on. And if anyone spots anything wrong on there, let me know!

• I answered elsewhere[1]. I’ll just note one important point from my reply: Asterisk and FreeSwitch are not PBXs. They are telephony servers. One application you can build using them is a PBX. You can either program it yourself or use an existing one (e.g.: FreePBX for Asterisk). It’s not clear from your question which of the two you need.

  To me personally the real advantage of open source is not the cost. It is the ability to tweak, and the control you retain. Right now you are new to VoIP. But that will soon change.

  [1] http://lists.pjsip.org/pipermail/pjsip_lists.pjsip.org/2017-April/019929.html

• The solution you choose should be based on many factors which should include your business requirements, team’s experience, your budget, growth expectations and more.

  You can choose Asterisk or Freeswitch as a platform and start building on that – but it is not simple and being new to VoIP you are likely to make mistakes. The “do-it-yourself” approach will some money initially, but will be the most expensive option long term – as you will be denying the economy of scale. Bringing a “smart programmer” won’t help much as you will also create a “lock-in”. In fact, this could be worse than a dependency created when you use a commercial or a known open source solution as while you would still be able to get help from the community for the “base” part of your pbx, your custom part will be much harder to deal with.

  Our company started building Asterisk based PBX in 2002 and Multi Tenant PBX in 2005 – we do this as our core business and are still finding areas for improvement :). As your experience with VoIP is minimal I would side with your CTO – you should find a solution high enough in the stack to avoid the complexity of building it all yourself.

  Good luck,

  Alex

• Thanks All.

  Thanks Alex, we also tested thirdlane PBX, and comparing it with PortSIP
  PBX, Vodia PBX, we hope we can make decision next week.

  Best regards,

• Have a look at xCally from Xenialabs too – they are particularly popular with call centers (and still asterisk based).

  From: asterisk-users-bounces@lists.digium.com [mailto:asterisk-users-bounces@lists.digium.com] The solution you choose should be based on many factors which should include your business requirements, team’s experience, your budget, growth expectations and more.

  You can choose Asterisk or Freeswitch as a platform and start building on that – but it is not simple and being new to VoIP you are likely to make mistakes. The “do-it-yourself” approach will some money initially, but will be the most expensive option long term – as you will be denying the economy of scale. Bringing a “smart programmer” won’t help much as you will also create a “lock-in”. In fact, this could be worse than a dependency created when you use a commercial or a known open source solution as while you would still be able to get help from the community for the “base” part of your pbx, your custom part will be much harder to deal with.

  Our company started building Asterisk based PBX in 2002 and Multi Tenant PBX in 2005 – we do this as our core business and are still finding areas for improvement :). As your experience with VoIP is minimal I would side with your CTO – you should find a solution high enough in the stack to avoid the complexity of building it all yourself.

  Good luck,

  Alex

• Well said Alex

• Hello,

  You can look Wazo, it’s a fork of XiVO and it’s a powerful graphical PBX with Asterisk.

• Hi all,

  Finally we make decision that go with PortSIP, the reasons are below:

  1. Support the easy cluster deployment for handle large concurrent calls and provide
  2. All REST API(this is very important to us for integrate the PBX with our current system), and also offer the rebrand app for free.
  3. The multi-tenant arch.

  Thanks all for your suggestions, we have learned a lot of !

  BR