New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
doc: Document distributed installations #271
Comments
Thanks cruzccl. I think what is most important is how to join a new server up to a cluster, as well as how to remove a profile from a machine that should not handle that task. The single server will likely remain the first database server, so adding a second database server, then moving media relays, portals, proxies etc off to other servers as load demands. |
About Portals, you talk about the platform administration portal, and we are right, it can works with an haproxy. |
In order to make this issue useful for everyone its goal will remain the same as it was described in the first note: improving both documentation and debian packages to make installing a distributed installation with one machine per profile as easy as possible. Migration from standalone installation to distributed installation won't be documented as we consider it the wrong way to go: standalone is meant for testing (with testing carriers and uacs) and distributed is production-ready. Scaling software different from Application Servers and Media relays (such as supervisor/gearmand) will remain out of the scope of this issue too. Sorry for this, but we need to focus our (limited) efforts as much as possible to make improvements that are useful for everyone. I hope you understand our point of view :) Regards, |
I totally agree the migration from standalone to distributed is the wrong way, testing in standalone is ok, but we must start with a distributed installation, even with a small one.
If you don't provide tutorials or ways to do it, you have to provide advices on interactions between every profiles, for us to understand and know how to build the best complete infrastructure. |
Sorry but I have to disagree in this fact. The software provided here is free (as in freedom) as we hope that the world can make the best use of it, or even request anyone who has the knowledge to update, improve and enhance it, but that doesn't come with an imposed requirement to us for giving support or documentation to anyone but our clients. We do our best documenting and giving free (as in money) support whenever we have the time, but that's not part of the project work: it's our contribution. Of course there is a lot of work to do, and a lot to improve, but we are a small team that has to focus on our client needs, using the time that is left to improve the documentation and processes so everyone benefit from our work. We don't have to provide free advice, support or documentation. We do it because we belive in free software and we care about whatever we do. That's where all our resposibility ends. We are very thankful to the users and community that tries our software and think it can help with any of their needs. We hope that community will help us to improve the software for everyone in form of contributions and/or paid requests. Regards! |
We may be able to help documenting some of this. We are doing some experimenting now. I think redis needs to be added to this list for artemis. Possibly by using sentinel. Also files in |
Do you have plan to try the nginx+php-fpm instead of the apache+php mode? |
Hi @learnin9 Please add questions as new issues, that way can be referenced in the future. The topic of this issue is documentation. We have no plan to schedule any resource to change our current apache configuration. Current one seems to work well. We are focused right now in backend and sip severs. Regards |
Hey @Kaian , we might have a consultancy gig for you guys. Where can we contact you? Thanks!! |
Hi @borland667 Check FAQ for contact information. For commercial support write to comercial@irontec.com Regards! |
In order to document distributed installation (asked in #269 ) package dependencies should be reviewed.
This way, installing IvozProvider with one independent machine for each profile should be as easy as installing corresponding packages and documenting manual fixes (currently made with Ansible in our installations).
Existing profiles are:
Scaling from one machine to multiple machines depends on the profile:
The only scaling process that will be explained in the official documentation is AS and media-relay scaling, as they do not need any additional software unrelated to IvozProvider.
The text was updated successfully, but these errors were encountered: