I installed Rockstor 4.5.5 (now updated to 4.5.6) on a Raspberry pi4. The boot drive is balanced. & scrubbed on each boot, adding 4 to 4.5 minutes to the boot time. My other Rockstor machine is a “real” computer* using SATA drives and doesn’t scrub & balance the system drive at boot time.
Question: is the scrub & balance at boot time expected on a machine booting from a micro SD card?
*With sincerest apologies to the Raspberry community.
@wdc Hello again, and thank for testing the Pi4 installer.
There is actually a 4.5.6-0 based installer also now.
There is some upstream (read openSUSE) automated maintenance of the system drive. We don’t enable anything ourselves on that front. And we have in-fact reduced some of our system drive tweaking of late. See:
So what you are seeing is non Rockstor specific but relates to openSUSE 15.4. Take a look in their forums etc for how this might be re-configured. And do report back on your findings as it may be something we would want to surface within the Web-UI when we get the human resources to effect those changes.
Maybe this is a real-time clock type thing going on. I.e. the upstream maintenance always thinks it the first boot or the same date on each boot. We do run ntpd or it’s kin thought!
That’s my guess. You can likely disable the auto scrub/balance and enable Rockstor managed ones if need be. But I’m not currently familiar with where these are instigated from. But there was a recent form post by @aremiaskfa on this matter where he noticed the upstream systemd btrfs-scrub.timer job that is presumably responsible for what you are seeing:
It may be that an upstream update fixes this anomaly of timing. Let us know what you find out on this one.