Yeah container's disk seem "fragile". There are a lot of corner cases just not covered that come up in the real world.
Mainly if ROSE or USB is not mounted when the container try to start, it doesn't try again. If disk name changes due plugging in new things, also a problem. And changing IP of VETH is generally problematic at best. If reboot was because of an RouterOS software upgrade...well...always possible change that break what was working.
My advice if this keep happening is not use "Start at Boot" option, and instead script the script in schedule to it starts after 30s or something (and checks the disk). And yeah use a quality USB disk and formatted by RouterOS might be a good idea too.
thanks, i was thinking of scripting the start anyways after a reboot. i use an internal ssd on the RB1100 for docker storage so this SHOULD (always) stay the same ;-D