I think that is lacking I find with rockstor is a simple and built in power management. I know I can set up cron from terminal, but normal user, this maybe trivial.
I as a non 24x7 user, would like to have the ability to power off my server at 11pm when everybody in the house is sleeping and to power back on automatically in the morning at 8am.
What do you think?
So far I am liking rockstor, I am trialling both rockstor and unraid, previously coming from OMV
This has been discussed before, and I have personnaly requested the same feature.
But as Rockstor is supposed to run in many different setups it would be hard to implement.
Its easy enough to make the system shut down at specific times, but the powerup would have to be implemented by the computer itself, e.g. by the system BIOS.Either via the user himself entering a startup time, or controllable from software (I know some Dell computers support this)
My system for one, does not have any settings for automatic startup at set times. And no way to control it from software. So such a feature would turn my computer off, and I would have to turn it on myself.
Rockstor advanced scheduled task is an open issue on Github
Notes about this:
my idea is to have it with some default available tasks (turno off, reboot, etc) + a complete user custom tasks handler
Note about shutdown: shutdown on an ext4 fs, all fine - shutdown on a btrfs fs while some scheduled tasks performing snapshots or scrubsâŠwell, we want to avoid this (someone could say âhey man, users schedule tasks so they have to care not to shutdown when tasks runningâ - not our idea, we must try to prevent user oversights)
Which has a reference to setting the bios rtc to wake a system (if the bios supports this of course).
This issue in turn links to:
And I think Mr @Flyer has already laid some ground work on this stuff.
@Flyer I donât see a problem with a clean shutdown of btrfs, balances will resume on next boot and I think scrubs just get cancelled. And given snapshots are essentially instantaneous they will be completed as part of a regular fs shutdown, if requested in time of course . But replacing a drive should be a blocker action, at least on our part I think; but an extreme example. But we donât yet have this in the UI anyway. Or were you thinking of side effects within our UI ie not robustly âtrackingâ scrub or balance jobs? Nice link to the ACPI stuff there, Iâve popped it in to the latter referenced issue above. Apologies if Iâm reading thing incorrectly here.