I believe (based on other posts) that there appears to be something going on specifically with the MR34 models, and the 25.11+ firmware versions. I would reach out to support and see if there are any known issues, the more people that do this, the more pressure they get that its not just a random few people with this issue.
You might also want to ask them to downgrade your MR34 to 24.12 (I believe this is the last code on 24 train that was stable). I think others have had positive feedback from doing this, and hopefully it will make things stable for you.
As for the reboots, I think it would be an amazing feature to have some sort of ability to schedule reboots within the dashboard, otherwise you'll have to create a script or use ansible or something complex.