Hey, sorry to say that I believe BalenaOS has gravely misconfigured the settings for this board…
Every now and then my unit will completely melt down when under load. This happens repeatably on two boards, this isn’t a problem in Armbian since they have VDD_CPUX set to 1.1v rather than 1.3v which resolves this issue, to my knowledge.
The “melt down” scenario happens repeatably when sending a multi-container project with balena push [ip]
This is a real let-down as I assumed Balena wouldn’t make these sort of mistakes, especially when the fix has been in Armbian for so long.
Is it possible that someone could figure out either how to set this using a supervisor envvar, or if it could be changed in a revision of the OS for the board as to prevent house fires? Thanks.
I’ve been informed this is not one of our supported boards, but is a community board (ie. maintained by the community at large). As such, it would be worth raising an issue (and PR!) on balena-allwinner repository. It’s been suggested that it may well be worth commenting on the following PR: https://github.com/balena-os/balena-allwinner/pull/65 which includes comments from the original maintainer.
Okay, whilst you’re here can I ask another question, as to why it’s not updated to the latest BalenaOS? If it’s all automated (as it is on the github link you’ve just shown me) why is it not being pushed to the Dashboard? Why are we stuck with 2.35+rev1, what determines whether you make it available to users or not?
Our release process is based on several factors. Although every release is indeed built internally, for all our supported boards we go through a heavy test cycle to ensure balenaOS releases are thoroughly ready for release.
For our community boards, we go through these at certain periods to update these device types for release. This ensures we don’t release broken images to our customers.