January 22, 200917 yr If a drive spins down from inactivity, the unRAID interface will show it as spun down. So far so good. If you spin it back up by reading SMART data, such as with smartctl, but do not read any data, the unRAID interface still thinks it is spun down. So it will never spin back down. No matter how many times you refresh the unRAID management interface, it still shows as spun down.
January 22, 200917 yr If a drive spins down from inactivity, the unRAID interface will show it as spun down. So far so good. If you spin it back up by reading SMART data, such as with smartctl, but do not read any data, the unRAID interface still thinks it is spun down. So it will never spin back down. No matter how many times you refresh the unRAID management interface, it still shows as spun down. It sounds like a flaw in the logic unRAID is currently using. In fact, it is the OTHER half of the issue discussed here where it will spin up a drive after we have spun it down purposely. I'd send a PM to Tom, noting this second issue with spin-up/spin-down. He not only needs to check if it is already sleeping before checking its temperature, but to also issue a sleep command, if it is spinning when it knows it has had no "I/O" activity for the time-out duration, if he finds it spinning. Joe L.
January 22, 200917 yr Nice find! In the myMain smart view, in order to try to avoid consecutive (lengthy) spinup of drives, I actually spin them up by accessing a random sector in parallel. This should mean that unRAID will see the access, realize that this disk is spun up, and spin it down in due time. But I think if you click on the "Sm" link in the info column, it will JUST run the smart report and will be subject to this issue. I'll put in a read access in there to prevent this bug. Note that since WD drives don't spin up with a smart request, this is not an issue for them.
Archived
This topic is now archived and is closed to further replies.