DjangoPBX end of year update

[Adrian Fretwell]

DjangoPBX has seen many developments and improvements over the last year. The project now has sponsorship from ITSPs in the UK, US, Germany and Turkey. Several ISPs are using DjangoPBX purely via its REST API to provide telephony functionality within their existing customer dashboards.

The main focus for this year has been developing the DjangoPBX Cluster. The Cluster is an innovative solution for tackling scalability and providing resilience whilst improving overall system performance. Details of what a cluster is, how it works, and how to build one, are given in the documentation link below:
https://www.djangopbx.com/static/documentation/cluster/cluster.html

During 2024 there have been 339 commits to the DjangoPBX repositories, which represents a large amount of development work. The Repositories can be found here:
https://codeberg.org/DjangoPBX

The DjangoPBX team would like to wish you all both the seasons greetings and a happy and prosperous VoIPing New Year!

 

[s2svoip]

unreal Adrian, this is really taking off, I really need to put some time back into a new install and run it for my own business & testing, love how well this is playing out. nice work! happy holidays

 

[gflow]

Wow fantastic documentation and project. I'll have to start testing this out.

 

[DigitalDaz]

Awesome Adrian, its come a long way.

Just a quick question: is the rabbitmq a single point of failure? I remember this being an issue with kazoo back in the early days.

 

[Adrian Fretwell]

The short answer is yes. In the diagram, I just show one rabbitmq for simplicity. The FreeSWITCH mod_amqp configuration does allow for a relatively simple secondary server to be used. RabbitMQ, does have it's own clustering arrangements for high availability, the details can be found here:
https://www.rabbitmq.com/docs/clustering

There are a few options, I have been running some tests using a warm standby rather than a secondary server. I'm still undecided as to the better way to go, I think it will depend largely on the size of the overall deployment.