Apologies for the newb question.
My preferred workflow has been to do a snapshot on the digitalocean dashboard before doing discourse updates, in case Bad Things Happen. As I am realizing now, just running
shutdown now -h appears to not be enough and shuts down the server while docker is running. Do I need to run this instead?
cd /var/discourse && ./launcher stop app && shutdown now -h
Is there a way to add this to a shutdown process so that the shutdown is clean? Or am I wrong in pinpointing this as the source of my woes? Is there another way my database could be getting corrupted?