Disks never spin down

[HellDiverUK]

Folks,

 

Running DrivePool with the following drives:  1xWD50EFRX, 1x WD80EFZX, 1xWD80EZZX.  OS is Server 2012R2.  Drivepool only handles TV shows, music and movies.

 

For some reason, even though the server isn't used much, it doesn't seem to spin down the drives.  Ever.  If I put the same data on a Storage Spaces simple pool, drives will spin down OK.

 

Any reason why DrivePool is forcing the drives to run 24/7?  Is there any way to log what's happening?

[Christopher (Drashna)]

Are you using the beta version (2.2.0.651 or higher) of StableBit DrivePool?

 

If so, do this:

http://wiki.covecube.com/StableBit_DrivePool_2.x_Advanced_Settings

Set "BitLocker_PoolPartUnlockDetect" to "False", and restart the system (or restart the "StableBit DrivePool Service"). 

 

If you're using the release version (2.1.1.561), then this shouldn't b e enabled unless you've done so manually. 

In this case, do you have StableBit Scanner installed? If so, the SMART queries can keep the drive active. Throttling the queries will fix this. This can be set in the "Scanner Settings", under the "SMART" tab. 

 

Also, check the frequency that Sonarr and Emby are scanning the drive.  Also, try turning off the real time detection for Emby, as this may frequently ping the pool. 

[baldursgate]

Are you using the beta version (2.2.0.651 or higher) of StableBit DrivePool?

 

If so, do this:

http://wiki.covecube.com/StableBit_DrivePool_2.x_Advanced_Settings

Set "BitLocker_PoolPartUnlockDetect" to "False", and restart the system (or restart the "StableBit DrivePool Service"). 

 

I'm using the 2.2 beta on Windows 8.1 with BitLocker, should I do this?

 

I just reinstalled the OS and so far BitLocker seems to be working fine with the 2.2 beta, but I haven't noticed if my disks spin down yet.

[Christopher (Drashna)]

I'm using the 2.2 beta on Windows 8.1 with BitLocker, should I do this?

 

I just reinstalled the OS and so far BitLocker seems to be working fine with the 2.2 beta, but I haven't noticed if my disks spin down yet.

If you're using BitLocker, and especially if you're not encrypting the system disk.... leave the option enabled. 

 

The issue is because we're querying the BitLocker status, and that seems to be "waking" the disks. 

 

Unfortunately, we don't really have a great solution for this. Though, if you are encrypting the system disk and using the automatic unlocking, try disabling it, as it *may* not be needed on your system. 

[greggerca]

I'm trying to figure out how to get Windows Server 2012 R2 Essentials to spin down my disks like WHS 2011 used to (both using DrivePool).

It may be an issue with the controller as I switched motherboards.

 

Are there any options / plug ins in DrivePool (besides the BitLocker one - that's False already) that might help?

Everything is basically "Default" on the release version of DrivePool.

 

I have Scanner installed right now, but they didn't spin down prior to its installation.

 

I installed Scanner actually to see if I could make it spin down disks proactively.

I'm thinking this pertains to the shabby controller on my Asus E35M-1 Pro.

 

I'm going to install an external JBOD array case and dedicated eSATA controller for it in a few days, so maybe this whole topic will be moot.

[Christopher (Drashna)]

Just a heads up, StableBit Scanner does query the drives very frequently.  In the Scanner settings, you can throttle the SMART queries, and this may help. 

 

 

Otherwise, check the power options in the Control Panel.  Make sure it's on balanced (not High Performance), and make sure that the disk idle time is set in the power plan. 

 

Also, try running "resmon" and keep an eye out for disk access.  Any access to the pool, or the underlying disks may keep the disks awake. 

 

Certain services, like WIndows Search may cause this to happen. 

[greggerca]

Thanks again... yeah, I know Windows Search was doing a lot of work after rebuilding the server.

I did set the power settings to be around 20 min or something for a spin down.

 

Once I remove this bad disk from the pool, maybe things will settle down.

 

I'm sure that running Plex Server, Windows Media Sharing, and Search all access the disks in unpredictable ways.

Once Scanner is done attempting to recover a bad file (or I cancel it), I can try SMART throttling. The disks weren't spinning down prior to Scanner being installed, but again, I bet Search was the culprit.

 

Thanks again!

[MetalGeek]

I'm having the same problem.  Windows 10 Pro,  Drivepool 2.2.0.651 and latest scanner beta with 5 WD 2TB Red drives:  1 identified as WD20EFRX-68AX9N0 and 4 as WD20EFRX-68EUZN0.  I made the suggested edit and restarted drivepool service, I've turned off drive indexing on the drive properties, disabled all backup software, and verified my drives are set to spindown after 30 minutes.  I even tried to  uninstall Scanner.  So far nothing changes the behavior.   Running resmon shows nothing touching anything but my C: drive. 

​

​

​What I am seeing is every 2-3 seconds there is a disc access of 4KB across all drives at the same time.  It's like a heart beat of some kind that doesn't show in resmon.  See attached screenshot from one of the drives and from scanner showing the drives getting hit all at once.

​

EDIT: Stopping the Windows search service had no effect either.

​

​

​

[Christopher (Drashna)]

The disks weren't spinning down prior to Scanner being installed

 

Stopping the search service may help.  I doubt it though. 

 

Basically, try stopping everything that may use the disk, including Plex.   See if you can find what exactly is triggering the activity. 

 

Or just ignore it and leave the disks awake.... (not the best solution, depending on your needs/wants). 

 

EDIT: Stopping the Windows search service had no effect either.

​

​Capture.PNG

​capture1.PNG

 

What controller are they hooked up to? 

 

I ask, because I had a HighPoint RocketRAID card that did this. Pinged the disk periodically.  It was doing so because the web management service was querying SMART info.  

[MetalGeek]

What controller are they hooked up to? 

 

I ask, because I had a HighPoint RocketRAID card that did this. Pinged the disk periodically.  It was doing so because the web management service was querying SMART info.  

 

​The integrated Intel controller. Intel 8 Series/C220 Chipset Family SATA AHCI Controller.

I know disks will spin down on the controller.  Used to have two WD RE 750GB drives and they spin down for years under Windows 7.  My Reds spin down fin on my Synology as well  I'm determining spin down success based on Power Draw off my Kill A Watt.  With Just an SSD my system pulls 38-40 watts, steady.  Tested this for many weeks.  With the 5 Drives Idle, power should be about 51 watts which is right in with what I measure of 48-52.  If the drives are asleep, I should see power in the 40-42 watt range but it never drops below 48.  Thta tells me the drive are idle not asleep, per WD drive specifications. 

 

I'll start testing various services/etc.  My first test will be to disable Drivepool/scanner and see if things change. 

​

[MetalGeek]

The heatbeat access is Drivepool.  When I stop the service the disc access stops.   Resmon does not show this disc access at all to any drive however scanner and task manager pick it up, both showng the same thing.  I'm going to leave Drivepool off all night and see if the drives are spun down in the AM. 

[Christopher (Drashna)]

The heatbeat access is Drivepool.  When I stop the service the disc access stops.   Resmon does not show this disc access at all to any drive however scanner and task manager pick it up, both showng the same thing.  I'm going to leave Drivepool off all night and see if the drives are spun down in the AM. 

 

 

Okay. 

 

If you're using the 2.2 beta (any build of it), make sure that the "BitLocker_PoolPartUnlockDetect" option is disabled in the advance config file:

http://wiki.covecube.com/StableBit_DrivePool_2.x_Advanced_Settings

 

It's enabled by default, and will cause this sort of activity. 

 

If it is disabled, you've rebooted the system and you're still seeing this, then enable file system logging, let this run for a bit, stop logging and upload the logs to us.

 http://wiki.covecube.com/StableBit_DrivePool_2.x_Log_Collection

[MetalGeek]

BitLocker_PoolPartUnlockDetect is off and has been for a week.  Logs incoming.  30 second capture.

[MetalGeek]

I did not open asupport ticket to go with the logs, should I have done this?

[Christopher (Drashna)]

I did not open asupport ticket to go with the logs, should I have done this?

 

Yes and no. 

 

It's not strictly required, and we will troubleshoot on the forums, but it's easier for us to do so in a ticket, and we prioritize them over forum posts.

 

And in some cases, we do end up require info that may be sensitive (such as logs).  So, again, we prefer the ticket system, since it allows for privacy. 

 

 

 

However, I've flagged the issue and logs for Alex. 

https://stablebit.com/Admin/IssueAnalysis/27161

[dannybuoy]

Related question - what's the easiest way to check if your drives are spun down or not? This is something I'm meaning to look into, once I have everything working as I want to, I want to look into power savings!

[Christopher (Drashna)]

Related question - what's the easiest way to check if your drives are spun down or not? This is something I'm meaning to look into, once I have everything working as I want to, I want to look into power savings!

 

that's a catch-22, unfortunately. 

 

From within Windows, querying that information is considered disk activity and will wake the disk. 

 

Sort of touching the drive and see if it's spinning .... which is NOT a good idea... the only other method is to measure power consumption.  Most drives consume 2-10W of power when active, and a lot less when idled.  So using your UPS or a Kill-A-Watt meter may be the best option. 

 

And if you're looking to cut power costs, then you should invest in either (both, really).  

 

 

Also, Alex (the Developer) has a good reply about power savings and some of what StableBit Scanner can do.

http://community.covecube.com/index.php?/topic/48-questions-regarding-hard-drive-spindownstandby/

[orangeblue]

Hi there,

 

Same problem here. Windows 10 professional fresh install, installed latest Drivepool Beta, now every 5 seconds it creates read IO on my drives, preventing it from spindown.

 

I've traced it back to an wmi query "Win32_EncryptableVolumeProvider"

 

Disabling "BitLocker_PoolPartUnlockDetect" did not solve the problem. Now i've installed the latest stable, which don't have the issue.

[Christopher (Drashna)]

Hi there,

 

Same problem here. Windows 10 professional fresh install, installed latest Drivepool Beta, now every 5 seconds it creates read IO on my drives, preventing it from spindown.

 

I've traced it back to an wmi query "Win32_EncryptableVolumeProvider"

 

Disabling "BitLocker_PoolPartUnlockDetect" did not solve the problem. Now i've installed the latest stable, which don't have the issue.

 

 

Then the issue is *definitely* the BitLocker setting. 

 

That's the main difference (settings wise) between the two. The beta's have this option enabled by default, where as the release version does not. 

 

Also, the "Win32_EncryptableVolumeProvider" call is being sent by the BitLocker status query. 

https://msdn.microsoft.com/en-us/library/windows/desktop/aa376483(v=vs.85).aspx

 

 

After making the settings change, you do need to reboot the system. 

Additionally, it may be best to use a 3rd party text editor in regards to the config file, as the built in "notepad" app has issues with the files. 

 

Also, make sure it's renamed to "DrivePool.Service.exe.config", and that you removed the "default" from the file name. 

[Christopher (Drashna)]

Could you try doing this: 

• Use FileSpy ( http://www.zezula.net/en/fstools/filespy.html) and attach to one of the volumes that is waking up.
• Wait for the wake up call and observe what process it is, and what is the command that's waking it up.

If it's WMI that's waking the disks, then it may be the BitLocker code that is causing this.