TN Core + apps server or TN Scale

[NinthWave]

I have been running Core since feb 2021.

The main purpose was of course the NAS part. But I soon realized I could move some servers that were taxing my old PC (2012) to TN using either jails or VM. Said apps are:

• Assetupnp server : a DLNA/UPnP music server
• Emby: a video server
• Seafile for file sync, just 2 clients

Then last month, I added a third IP camera to my house and soon realized that movement detection (without a 25$ Google Coral now sold >350U$ used due worldwide unavailability) was way more taxing than all other apps I moved to TN. So now I want to use Frigate.

A friend, which happen to teach informatics, told me one should avoid put all his eggs in one basket and keep the NAS just for this single purpose and move all jails and VM (only 1) to another dedicated piece of equipment. So he gave me a desktop Intel server (E3-1230v3) and I have started fiddling with Docker/Portainer.

So here are my questionnings:

• While the Dell T610 draws 150W, it is just idling because it's more powerfull than the tasks. Running the Docker server will add 60W plus use some space.
• I understand that ideally one should avoid the single basket all eggs situation but on the other hand, a lot of person here are using plugins and jails in Core so it is viable.
• From what I have read, I am not sure Frigate will work on Core.

So I am seeking any arguments in favor of:

• Stop keeping a very capable T610 idling and pursue the VM and jail way in it, still with Core.
• Keep Core just as the NAS and migrate Asset, Emby, Frigate and any future projects in the Intel Docker server
• Keep the T610 at work but migrate from Core to Scale

I am listening, thanks to all.

 

[sretalla]

one should avoid put all his eggs in one basket and keep the NAS just for this single purpose and move all jails and VM (only 1) to another dedicated piece of equipment

That argument only applies if there's some kind of performance or risk penalty you will pay by keeping everything in one box.

If you're happy with the cost and performance and risk of the solution(s) all becoming unavailable at once, but only consuming the resources/power of one single box, then where's the problem?

As long as you have a strategy for backing up your important data and the config of each of your apps, everything can be put back in a relatively small amount of time following some kind of major failure.