I was wondering about booting from a DELL dual SD card configuration (IDSDM). I will be moving my unraid setup that has been going for about 10yrs to new hardware in the next year or so. I'm doing some planning at the moment on how I want it setup. I recently acquired an R630 with the dual sd card configuration inside.

 

These cards are mirrored (can be changed). There is a primary card that is booted from (slot 1) and then the secondary card is mirrored if anything on the primary card is changed. At boot, they are checked against each other and if anything is different then it will ask you if you'd like to continue to boot of slot 1 or if you'd like to restore card 1 from card 2.

 

I think the reason for wanting to boot in this manner would be to retain a "backup" at all times of your "USB drive", which I've always thought of as an achilles heal because if it fails then you lose very important information regarding the server. I could be wrong, but it seems like you would need a backup of it every moment in order to recover from a failure.

 

So... why could this work? Do SD cards not have a GUID that could be registered for the license? Since the second card wouldn't be registered then would there be write issues to the second card for redundancy? The secondary card is never booted from, so I would think the GUID wouldn't matter in that sense. If the first card were to fail then you could get a new card, register it with Unraid Support and then completely recover the exact configuration before the primary card failed from the secondary card.

 

To me, that seems very advantageous. I'm sure I am missing something here, but I'd love to be enlightened so I can either rule this out or consider a way forward for my move to the new hardware.

I think it’s kinda like the way RaspberryPi have moved on from SD booting to USB due to wrecking cards. 
 

I wouldn’t want that Achilles heel. 
 

to backup your flash drive. There is an auto backup you can with a plugin. 

Edited July 1, 20215 yr by Jammy B

Unraid makes very few writes to the boot device as it runs in RAM. It's unlikely you'd get failures in either device, but I would rather have a dual sd card configuration because you have a mirrored copy of your boot device and the config files etc.... This is something very common in the enterprise industry. ESXi is run this way a lot of times on these servers.

The only supported boot method is from USB flash drive as the licence is linked to the flash drive's GUID.

Thanks. That answers that question.