Well, windows 10 can be installed on a USB but you need to google it.
If you just want the Magic Mirror application on a UsB stick then put it in a folder named MagicMirror on your stick and adapt the command line/batch script accordingly.
J

If you have a Windows 10 computer, why would you need a USB drive?

Which Raspberry Pi do you have? The Raspberry Pi 3 can boot directly from USB but the others need a SD card.

Download the chosen image and write it to UsB in the same way you created the SD cards. Then remove and insert the USB stick from your computee and copy the file bootcode.bin in the /boot directory to an empty SD card. (This card is needed unless you have a Pi3 but it will only be read from so no signifikant wear.)

Insert the USB stick and boot your Pi. That should do it.

If you have a Pi3 there is a different procedure, google is your friend.

Yes, especially if you put the /var lib on a tmpfs which resides in ram.

I am running both my mirrors from usb.
The trick is to have a SD Card with only bootcode.bin on it. That is the boot loader which will try to load firmware from attached devices.

Download raspbian or whatever image you want and write it to the usb drive with etcher or similar.

You need to make sure that cmdlint.txt anf /etc/fstab point to the usb drive but that should alteady be the case.

@BananaBro just use the same region name, e.g. “bottom_right”, for both modules and they will be automatically stacked vertically.

@j-lewis,
For some reason your config.js file is missing or inaccessible. Can you issue the command
ls -l /home/ubuntu/MagicMirror/config
and post the output. It would also be easier to help if you can post the content of the file /home/ubuntu/MagicMirror/config/config.js.

@dsmith76
If I remember correctly the calendar needs to be public and not solely shared with users.

I agree with others that the problem is most likely related to Electron and not module-specific but rather related to which module dependencies it has.

@retroflex
I will dig up another respberry pi and try it out. My test environment is a dietpi vm which runs on a pc so I wouldn’t expect the cpu load to be representative of the load on the pi.