Hi all, Can you offer some wisdom/suggest the art of the possible?
I have a small server setup i primarily use as a plex/SMB share server, and occasionally dabble in the occasional self hosted VM or containers.
My journey so far:
I was originally running truenas on an aging HP gen 8 microserver.
It was running a basic but fairly large SMB share and a plex container.
I've recently used a gen 10 intel NUC as a proxmox host and have successfully set up plex and turnkey within LXC containers and it works well with hardware transcoding etc.
Disk images etc are stored currently on the local drive of the NUC. At the moment my primary datastore is still running on the truenas/microserver and the smb share is mounted in to proxmox to pass through to the LXC containers.
I purchased a 6 bay teramaster DAS box which is attached to the nuc via USB3.2 and created a test ZFS pool to present in turnkey etc and this works fine/passes smart data etc.
My planned next phase was to purchase bigger drives for the DAS box and migrate the data from my old server. The old server could then be moved to another location and used for backups if desired.
The area id like advice is... I do have some other NUC devices, and would like to experiment with HA clustering (having no prior experience) for select LXC containers.
Given that I've gone down the DAS route how would i handle migrating machines that have storage and gpu passthrough enabled?
I would likely use a 2nd NUC as the 2nd node, I think ideally i would purchase a 2nd matching DAS box and set up ZFS replication between the two.
If I initially wanted to avoid this expense would it be possible to use the micro server in a similar way? I would have no issue ditching truenas for proxmox but id still want a nuc hosting the LXC containers for GPU passthrough etc.
any advice appreciated.
Cheers
I have a small server setup i primarily use as a plex/SMB share server, and occasionally dabble in the occasional self hosted VM or containers.
My journey so far:
I was originally running truenas on an aging HP gen 8 microserver.
It was running a basic but fairly large SMB share and a plex container.
I've recently used a gen 10 intel NUC as a proxmox host and have successfully set up plex and turnkey within LXC containers and it works well with hardware transcoding etc.
Disk images etc are stored currently on the local drive of the NUC. At the moment my primary datastore is still running on the truenas/microserver and the smb share is mounted in to proxmox to pass through to the LXC containers.
I purchased a 6 bay teramaster DAS box which is attached to the nuc via USB3.2 and created a test ZFS pool to present in turnkey etc and this works fine/passes smart data etc.
My planned next phase was to purchase bigger drives for the DAS box and migrate the data from my old server. The old server could then be moved to another location and used for backups if desired.
The area id like advice is... I do have some other NUC devices, and would like to experiment with HA clustering (having no prior experience) for select LXC containers.
Given that I've gone down the DAS route how would i handle migrating machines that have storage and gpu passthrough enabled?
I would likely use a 2nd NUC as the 2nd node, I think ideally i would purchase a 2nd matching DAS box and set up ZFS replication between the two.
If I initially wanted to avoid this expense would it be possible to use the micro server in a similar way? I would have no issue ditching truenas for proxmox but id still want a nuc hosting the LXC containers for GPU passthrough etc.
any advice appreciated.
Cheers