So I understand that VMs are best practice for running Docker Swarm in a cluster. But I want to pass my GPUs though so I can run some local LLMs.
I've managed to get docker running on debian LCX, but cant get to portainer seems like a common networking issue.
I've also had issues in the past of GPU passthrough breaking after an update with VMs
So my question is, as someone that just wants it to work and is time poor. What would be the best way, try to get the LCX working or put up with potential breaks with GPU pass though
Also I feel that using containers will be more efficient, but that's just an intuition, I don't know if i'm thinking about that right.
Seems i'm damned either way really HAHA
Thanks guys
I've managed to get docker running on debian LCX, but cant get to portainer seems like a common networking issue.
I've also had issues in the past of GPU passthrough breaking after an update with VMs
So my question is, as someone that just wants it to work and is time poor. What would be the best way, try to get the LCX working or put up with potential breaks with GPU pass though
Also I feel that using containers will be more efficient, but that's just an intuition, I don't know if i'm thinking about that right.
Seems i'm damned either way really HAHA
Thanks guys