Code Change Request

# 12410

Back to Code Changes

Christopher
Technical Support
StableBit Scanner
2.5.1.3062
Windows Server 2012 R2
Public
Alex

* [Issue #12410] Consider SAT a safe pass-through method for non-USB devices.

Turns out nothing was wrong with our stand by, we were simply not using SAT unless Unsafe Direct I/O was enabled.

I recall that SAT used to be a problem for some controllers, but times have moved on and more and more controllers are now supporting it. I've added it to the "safe" list of methods and it should now work in the case where smartctl -d sat was working.
Public
Alex

Investigate why stand by is not working in the latest builds.
Public
Alex

Then apparently, HDDScan is using a spin down command which puts the drive into a state where it doesn't wake up on the next command (which is not what we do). Given that sending a spin down command to the drive directly, inherently doesn't inform the OS, this is actually dangerous, as the OS can try to send a critical command to the drive and error out (such as saving transactional file system metadata). Although, it should retry the failure a number of times.

As for Scanner vs. smartctl (I presume they're sending the same command as us, but haven't looked at their code), there might be something wrong with the Scanner's stand by in general. I think something may be broken in the latest BETA (and I'm not how far back it goes.

I'm going to look into this soon.
Public
Alex

The standby option send the ATA STANDBY (0xE2) command to the disk using Direct I/O, effectively causing the disk to spin down immediately. The drive's board remains active and powered up waiting for the next command to come in and will spin the drive up when it does.

Not sure what this wiki is, here's a description I found by Googling:
http://wiki.osdev.org/ATA/ATAPI_Power_Management