Re: X11 forwarding with headless Ubuntu host with podman
nwe <[email protected]>
| Newsgroups | gmane.linux.debian.user |
|---|---|
| Message-ID | <[email protected]> |
On 5/25/26 2:52 AM, Christian Gelinek wrote: > > I have some software with X11 GUI I'd like to run inside podman > containers on machine octopus. The octopus host itself should be as > minimal and as disposable as practical and only be used as a gateway > to the software in the containers. Previously it was enough to have a > few real metal PCs (like marlin) with a full-blown desktop environment > but recently the number of OS variants has been increasing and I'd > like to save resources by using containers instead, where I'm hoping > that trying & installing new OS distributions and versions would be > less hassle and costly than dealing with real metal for each combination. > Not sure if the following would serve your purpose; it probably takes more resources than your attempted method. What I've been doing: Headless server running virsh virtual machines. The hardware is a 10 year old dual-CPU rack server with 256GB RAM and ~30TB hard drive pool so all kinds of capacity to play with. I used boxes and virtualbox al long time ago but those had various problems, I don't recall all the details, I think boxes was randomly crashing for me and virtualbox kept fighting my preference of monitor size in the vm. With both of those I was running the server with a desktop environment. Virsh I administrate entirely via ssh to the headless host. I run virtual machines: Windows 10, multiple versions of Ubuntu and Debian, and a couple other odd distros, all at once, and connect to their desktop via vnc over my internal lan. I've learned to start including the vnc port in the vm name to help myself remember for example, that vm ReactOs5915 has its gui on port 5915. A server with much less resources should handle the same virtual machines if you only run one or a few at a time. As long as the cpu supports virtualization, I've run a vm on a host with 4GB RAM. I used to think virtual machines take a huge performance hit compared to real hardware, but that seems to have significantly improved in recent years.
OpenPGP_signature.asc
(application/pgp-signature, 665 B)
-----BEGIN PGP SIGNATURE----- wsD5BAABCAAjFiEEk1oEzw7NPoezBiPm+kcZs1LgZ50FAmoUSAsFAwAAAAAACgkQ+kcZs1LgZ530 DQv/Tm8ja5f+Xz+ADjONecng+wAs4OGn6WOSw4b83EvmuHT702lWFALCGGKA7915cvEogE1Y3rQi yT/qZX7nWXAVVRxO1vcQykHLV3t3COpLe3acXnNVSgCRZkN9gCfKtqnj/5ei1jPIgEDjqMEM8QPe v3hnZwIuAtCUinbK/+3qo501ex+9Xnrd1PdvqaxvW/16+MkqS15vVS9Wrs8lU8EOYQhkXgzOtqQy BHf0qym0P+Yt6270+XrPGym1H+ykQ2bivzKT47EHAxamCuehMB3M9ZXTMrCDjJtw10b0jlQSdL3r teAGr5/K4cC6ud6HzkBZh1wdrVvuf9Y38kZ+OfHpOSkhqqb+Nn7Nv/64M7aYY636wzoEW0Drmbzc KFyvCtXuq2xRVi0itmgDJsjAWJBTgtyMFcPHtdAE35EMPKUxFQsp+dP3PQaguMe4ch8ucNu+pQDk veSOK5yLgYCqMCU9cvMRleZTZi04hrvtyG5KBwgIX3f4UrWpHrk0/H/s0T1M =tPaZ -----END PGP SIGNATURE-----