Jump to content

Mooseable

Members
  • Posts

    212
  • Joined

  • Days Won

    4

Posts posted by Mooseable

  1. I hope this is allowed and seen as a positive movement. If the mods don't want this here, feel free to remove it.

    With the delayed release of v5, a group of hard working individuals have gathered to update Kazoo v4.3 to support newer operating systems. We are in the final stretches, in the testing and documentation phase. Once the testing and documentation is complete, I'll go back to developing educational content and adding functionality to the platform.

    I still hope for a release of v5, as I'd rather put this communities efforts towards improving that over the older v4.3. If 2600hz would like to help us in any way, even if it's to answer some questions about some of Kazoo's functionality in the open source version, please reach out.

    Info on the repos, discord, or if you want to help contribute, can be found here: https://kazoo-classic.github.io/

  2. For anyone in the same situation, wanting to continue to run 4.3 but on a supported OS, it's been hard forked to https://github.com/kazoo-classic as I'm going to assume 2600hz/ooma won't want to maintain or accept PRs for 4.3.

    Repos will be updated as we test fresh deployments of 4.3 with installation instructions.

    That said, I'd rather put efforts into contributing to v5, but its closed source so I'm doing what I can :)

     

  3. As a quick update, we have Kazoo 4.3 running on a mix of Debian 11, Debian 12 and Alma/Rocky 8. Haven't yet stress tested it, but we can at least place internal calls (not yet connected with CTS trunks), nor have all apps/features been tested.

    @mc_ - are the rpms for Rocky8/9 for Kamailio/Freeswitch for Kazoo v5 only?


    For everyone else, here's the "cliff notes" on versions needed to run 4.3. Once the installations are stress tested, I will work on documentation that others can follow.

    🛋️  CouchDB [Version: 3.2.3] [OS: Debian 12] 
    ☑️ Running and working. Requires special HAPROXY config to reroute port 5986

    🛜  Freeswitch [Version: 1.10.9] [OS: Debian 11] 
    ☑️ Running and working. Anything later than 1.10.9 fails to work with legacy messaging from kazoo-applications on the AMQP bus and crashes freeswitch.  Requires a signalwire personal access token to access the repo

    🔐  Kamailio [Version: 5.5.7] [OS: Debian 11] 
    ☑️ Running and with a phone registered. Requires use of PostgreSQL instead of KazooDB. Config exists thanks to a fork from ruhnet.

    ⚙️  Kazoo-Applications [Version: 4.3 with OTP 19.3] [OS: Alma Linux 8] 
    ☑️ Running and working. Requires separate RabbitMQ node/docker image. 

    🐇  RabbitMQ  [Version: 3.13.7] [OS: Docker] 
    ☑️ No notes other than its best to be run inside a docker container on the same node as your kazoo-applications node.

    👾  Monster-UI [Version: 4.3] [OS: Alma Linux 9] 
    Working. Had to "compile" it on an older machine, then just tarballed the result and copied to Alma 9 due to issues with node package versions and GULP. Since this only needs to run a web server, you can run this on any OS.

  4. So Kazoo-Applications and RabbitMQ builds can be done on Rocky/Alma 8. On 9, I can't get Erlang OTP 19.3 working, and kazoo won't build with Erlang OTP 20.3 or higher.
    I have a freeswitch node up using Debian12, but I haven't tried sending calls to it yet as I need to get Kamailio working first.

    If anyone wants to collab on the efforts to build on Rocky/Alma 8 or 9, let me know.

  5. On 6/8/2024 at 3:52 AM, KNERD said:

    It was already mentioned the will be using Alma Linux

    What I meant was, alma 9 will be end of support in under 3 years. If it's 6 years before a distro update let alone a version update available, running kazoo open source is a massive risk.

    Only way to go is commercial but then, I never would have even heard of Kazoo if not for the open source community version.

  6. Even IF v5 comes out, you now know you'll have to run EOL operating systems in the future, when the v5 branch stagnates for 6+ years.

    I love kazoo, I submitted bug fixes (which never were approved because work was being done on v5) created community content, videos, community support.

    I even tried to buy commercial Kazoo but we are "too small" for it to be viable financially (under 1000 extensions) in today's competitive market.

    I'd love to continue with open source until we hit the threshold where I can hand off the system management to a commercial contract.

    But, it is what it is. I thank the kazoo Devs for building a fantastic system and I wish v5 was out before CentOS EOL

  7. Ooma built their solution on 2600hz open source. They also acquired OnSiP

    On the investors call;

    They see 2600hz as the solution for the future. They want to invest more into development of the platform core and applications.

    There are significant savings in development for Ooma through the acquisition (2M/year).

    The recognize the revenue model of 2600hz, and list the open source market as an opportunity, though how was not discussed.
    They also mentioned that the GP% of 2600hz is lower than Ooma, and they hope to increase it to match Ooma's ~60% GP within 6 months (or at least EBIT positive)
    When asked about the developer community, its size and opportunities there, they mentioned that there "is a community" but was not really elaborated on.
    The acquisition is largely around the technology stack and the operational benefits it brings, control over the roadmap, new opportunities to reach large carriers, and not just the existing customer base.

    in closing remarks, it was said by Eric Stang that they are committed to the 2600hz community and the open source version

  8. @mc_ appreciate the input. I'm personally happy to pay 2600hz, especially for apps. However, a small operation like mine tried to get a quote and it was all USA hosting for a lot considering our small customer base.

    I did suggest that 2600hz could host the kazoo apps nodes only, keeping the IP protected without having to run a full cluster, which seemed to garner some positive feedback from 2600hz, but I don't think it went anywhere.

    Its probably what the app exchange is targeted for.

    I'm happy to contribute to docs, for v5. But it has felt like a wasted effort to write docs for v4 since it seems the change to v5 would be drastic.

    I personally would love to hear what would be most helpful to contribute. I'm only one person and I'm extremely time poor, so knowing where my efforts would be most impactful would be good to know.

    I will add, I do try to contribute from time to time, but my PR from 3 years ago is still open (and still fixes an active issue)

  9. I also get why they don't want to release v5 as per a previous statement that people tend to "jump right to it" without testing themselves, then probably cause a bunch of complaints/support requests.

    But if we are anywhere near an open-source release, I would love if some of us could gain access to or sign up for the v5 open-source release for testing.

    But as @Chris Labonne stated, we are in that process of deciding to stick with Kazoo or not. And I'd love to do another Kazoo Community Training video with the updated v5. I've kinda held off creating anything as I don't want it to be outdated.

×
×
  • Create New...