I hate browser vendors/projects for that. Seriously. What about a perfectly working printer, UPS, switch ... with a web management that only supports SSLv3? This is my private network. I know what I am doing. But of course this is outside of your and my sphere of influence. For all these devices I have come to the conclusion that it's probably better to connect to them unencrypted, because support for plain HTTP is not completely going away but obsolete SSL standards definitely are. So effectively browsers are lowering security because they force me back to plain text for certain applications.
I was refering mostly to Traefik, when it comes what we support.
If you use nodePorts you're on your own when it comes to SSL or not, thats outside our sphere of influence.
Currently I have it running in a VM and ports 80 and 443 mapped to the host IP. So I can reach TrueCommand at
http://truecommand.local/ or
http://192.168.1.20/. If I understand you correctly, either distinct ports per application or proper ingres with a real FQDN are mandatory with Scale apps. Makes sense for public stuff, I'm not complaining. OK ... a little bit.
Actually, fake FQDN with selfsigned certificates are definately possible, but unavailable due to a UI bug.
Bug fix is already made, would take a few days to reach the actuall app, but afterwards you could select "none" for CertManager Provider and just use fake FQDN's for every app with "selfsigned" certificates.
I don't consider mapping container A to host IP 1 and container B to host IP 2 with no ingress, no SSL, no nothing, "advanced" but rather the most basic setup imaginable.
Firstoff: stop using the word containers, it's confusing.
SCALE apps are NOT containers, they are Apps based on Helm charts, which are based on K8S. A "container" is so many layers of abstraction below that, that will only confuse people.
Specifying specific interfaces for nodePorts is NOT easy and not something I even know how to do in such a way I can write a SCALE UI for it.
Feel free to send a PR if you know beter.
Sure. Thank you very much, I just needed some more hints to get the big picture. Looks like I'll stick with one VM per Docker app for the moment for production and not move these over to Scale right now.
More hints on what?!
At least send screenshot of what config you used, or name the settings you have doubts about.
SCALE Apps are not a docker container, not a helm chart, not a k8s deploytment... they are SCALE Apps.
The UI is it's own thing, it's custom crafted for every App.
So unless you tell me what you want to know, how can I help you? A bigger picture of the UI? Most options already have help-text, that help-text IS the big picture.
I really want to help you out here, but as long as you keep being vague about what you want to know, I can't help you.
TLDR: Ask the god damn question.
Because it looked like ingress is the only way to run multiple applications on the same port.
I think you always need reverse proxies to run multiple applications on the same port. (a single port, means a single port on a single network interface).
Thanks again, don't take this as criticism, please. It's just me thinking out loud and coming to terms with these new concepts.
Patrick
No worries, it's just that I really want to help out but you haven't given me anything to go with...