SSD never going to sleep

[ej24]

I have separate drives for Windows 10 and High Sierra. While checking Disk Info i noticed my Windows ssd has about 1000 hours on it over 2 years. Not bad. Windows is set to put drives to sleep after 5 minutes, so that sounds about right. But my Mac OS ssd has 12,500 powered on hours over 2 years. My usage is maybe 70/30 mostly Mac OS. So it seems my drive hasn't gone to sleep at all. Is this normal for Mac OS? My Mac OS ssd is a cheap OEM TLC based Toshiba drive so it's not the most robust drive. I'd like it to go to sleep whenever possible. How do I get it to go to sleep?

 

[BreBo]

How do I get it to go to sleep?

System Preferences > Energy Saver > Put hard disks to sleep when possible

 

[ej24]

I suppose I should have specified, other than that. That option is checked but the disk is never going to sleep judging by the powered on hours. Total writes aren't too bad but drive life remaining is down to like 80%. So I'd like the drive to go to sleep as much as possible.

 

[ej24]

Is there a way to manually change the disk sleep timeout via the terminal? Maybe I can manually change it to a shorter time to see if that forces the drive to sleep? Or could this just be an eccentricity of this particular Toshiba drive/controller in combination with Mac OS?

 

[UtterDisbelief]

Is there a way to manually change the disk sleep timeout via the terminal? Maybe I can manually change it to a shorter time to see if that forces the drive to sleep? Or could this just be an eccentricity of this particular Toshiba drive/controller in combination with Mac OS?

Hello there.

I'm not sure the same wear characteristics apply to an SSD when compared with an HDD. There are no moving parts in an SSD and the reason for a HDD spinning down is to save power and reduce mechanical wear. For an SSD the only wear that matters is Write cycles. Once you get down to zero %, theory has it you can still read all your data but no-longer write. So it make sense to reduce writing to disk when not necessary.

I know you don't mention it specifically but I feel a suggested comparison between the wear characteristics of an HDD with an SSD is confusing as they are two different approaches to data storage.

To address the problem of power-on hours or sleep, I would think that the drive controller of an HDD is still operational when the drive is "sleeping". Same for an SSD. However what is causing the constant macOS write cycles and thus causing wear? Well perhaps that is Trim and Garbage Collection of SSD firmware in comparison with HDD disk optimisation and defrag.

Quite why Windows reports a twelfth of the activity of macOS is more difficult to pin down. Could it be that macOS is being more accurate or reporting a different operation? Your very low figure seems to imply that Windows SSD activity is as infrequent as 1.37 hours per day total on average. Does that sound right?

I have no definitive answer, just my take on it. Others may shed more light. Interesting question though

 

[ej24]

Hello there.

I'm not sure the same wear characteristics apply to an SSD when compared with an HDD. There are no moving parts in an SSD and the reason for a HDD spinning down is to save power and reduce mechanical wear. For an SSD the only wear that matters is Write cycles. Once you get down to zero %, theory has it you can still read all your data but no-longer write. So it make sense to reduce writing to disk when not necessary.

I know you don't mention it specifically but I feel a suggested comparison between the wear characteristics of an HDD with an SSD is confusing as they are two different approaches to data storage.

To address the problem of power-on hours or sleep, I would think that the drive controller of an HDD is still operational when the drive is "sleeping". Same for an SSD. However what is causing the constant macOS write cycles and thus causing wear? Well perhaps that is Trim and Garbage Collection of SSD firmware in comparison with HDD disk optimisation and defrag.

Quite why Windows reports a twelfth of the activity of macOS is more difficult to pin down. Could it be that macOS is being more accurate or reporting a different operation? Your very low figure seems to imply that Windows SSD activity is as infrequent as 1.37 hours per day total on average. Does that sound right?

I have no definitive answer, just my take on it. Others may shed more light. Interesting question though

I realize "wear" is totally different for SSDs and HDDs. In my experience with windows, if you don't make read or write requests the drive controller puts the drive to sleep. So while many of my windows systems have power on times of 1000's of hours the drives typically have much less (maybe 1/10th) power on time because they're put to sleep (thus not reading or writing or P/E cycles). My main concern is why Mac OS isn't putting the drive to sleep, ever. To me, that implies the OS is making read and write requests, albeit tiny ones, all the time. A few kb ever other second adds up to many P/E cycles over time when the system is left running almost 24/7 in a work environment.

Also the operating system shouldn't make a difference in reporting power on hours. That's pulled directly from the drive controller. From what I can tell, Mac OS doesn't appear to be putting the drive to sleep, ever. Surely in actual macs like Macbook Pros the drives go to sleep whenever possible to save power.

So I guess I'm not really concerned about "power on" hours so much as, why isn't the drive sleeping? It seems like it's because the OS is constantly reading or writing things. It has about 156 total drive writes (about 40TB written on a 256gb ssd), which is why I'm wondering if Mac OS is making a ton of tiny, but constant r/w requests. Over the 12000 hours that comes to 0.3 drive writes per day. That's pretty high for a consumer drive, especially a cheap repurposed OEM Toshiba drive.