Running xpra in a pod/container

[louis-mulder]

referencing to r28061
I am not a experienced Python programmer so I attach my pamexec_auth.py module in Python version 2. This module works with the help of some bash scripts to startup a pod with a xpra-session. The Xpra proxy.... runs as an ingress-server in a K8 cluster. I have seen the new interface however after creating the pod it must update the capability display so the proxy knows where it must connect.
How difficult is it to convert this module to Python3 and using the new interface ?

Many thanks on the forehand.....
Louis

[louis-mulder]

Python version of pamexec_auth.py

[louis-mulder]

Bash script to startup a pod or get the ip-info

[louis-mulder]

bash startup script xpra proxyand freeipa client

[louis-mulder]

deploying a Xpra ingress-server, shared storage is based on NFS

[Antoine Martin]

So this is a follow up to #2933, right? Or a duplicate?
The ticket title is exactly the same.

I seem to remember saying this before but I can't locate the ticket comment: you should split your pamexec module: there is no need to duplicate so much code.
(and even if there was, you should import most of it from the other modules, not duplicate it)
Just configure your socket with two authentication modules (see wiki/Authentication), the first one is pam and the second one can do what you want.

See also ticket:2933#comment:2 wrt capabilities. Don't use a global.

The command should be configurable.

Don't do this:

argv = c.strget("argv")

Just do: argv = c.strtupleget("argv") then you don't have to parse strings.

display = my_argv.split(u"/")[-1]

If you want to know the display specified by the user, there is a much easier way IIRC.

[migration script]

this ticket has been moved to: ​https://github.com/Xpra-org/xpra/issues/2989