HD power down

[HighPerf]

does HDS prevent my HD from powering down? i have windows power management set to turn of my secondary drive after 20 min but the drive temp stays the same and i don't hear it spin up.

[hdsentinel]

Yes. When HDSentinel detects disk information, the drive resets the idle timer. By default HDS queries disk information every 5 minutes, so if the "turn off drive" option is set to 20 minutes, the drive cannot spin down.
Please open Configuration -> Advanced Options page and set Detection frequency to a setting which is larger than the power management timeout (for example, set to 30 minutes or more).
If Hard Disk Sentinel detects that the drive is sleeping, it does not wake up to update information (and the drive "Information" page will show that "Power state" is "Sleeping").

does HDS prevent my HD from powering down? i have windows power management set to turn of my secondary drive after 20 min but the drive temp stays the same and i don't hear it spin up.

[HighPerf]

thank you for the fast response. i did what you suggested and i think it worked but now i changed my storage drive and it does not work like that any more? i am using a WD drive as storage (full details below) and have windows spin down the drive after 20 min and HDS detection frequency set to 30 min. the drive will spin down but every 30 min it will spin back up. if i close HDS the drive stays off?

i don't need temp status of my storage drive and have disabled the tray icon for that drive. is there a way of disabling the temp monitor for just that drive? i would like HDS to keep the log info but i don't need the temp.

Western Digital Caviar Black WD1002FAEX 1TB 7200 RPM 64MB Cache SATA 6.0Gb/s 3.5" Internal Hard Drive

[hdsentinel]

If the drive spins up again, it means the operating system do not provide accurate information for Hard Disk Sentinel when it detects the power status of the drive (active or sleeping). Because the operating system does not provide "the drive is sleeping", Hard Disk Sentinel tries the detection (not only temperature but also other parameters) - which operation wakes up the drive.

Can you please use the Report menu Send test report to developer option? This way I can verify the hard disk controller and current driver version which can be the source of this problem. If possible, please check if there is a different (newer) driver for the disk controller as it may work differently.

[VGER]

Hi,

I'm currently working the same problem. I usually set the HDD spindown time considerably high (like 60 minutes) because I don't want them to continually spin up/down, but only to go down at night or when there is no access over prolonged periods of time.

Now I'd need to set HD-Sentinel to check even less often, e.g. 70 minutes. However, this defeats the whole purpose of the program as an early-warning system not only for SMART issues, but also for temperature alarms due to failed fans.

Does any information request from HDS reset the drive's timer? For example, it might not be necessary to check SMART data every five minutes, but temperature should be updated more often. Would it be possible to group the requests into "timer-safe" and "timer-reset" groups, that could be timed independently?

Or could there be a "timetable" for the checks? Maybe one could specify a "daytime interval", for example from 9 AM to 5 PM. Inside this interval, checks would be made with frequency A. Outside, checks would be made with frequency B. Of course, for you as developer this would be a lot of trouble for just this one situation.

One other thing: When HDS is started, will it cause sleeping drives to spin up, or will the power-down status be preserved?

Regards,
VGER

[VGER]

One afterthought on the "timetable check idea": How about dynamic?

"If the value didn't change for the last x checks, do not check for another y intervals, but force check at least every z intervals"

This would mean that if e.g. the temperature of a drive hasn't changed for 20 minutes (with interval set to 5 minutes), than it wouldn't be checked for another 20 minutes. After 60 minutes, however, a check will be performed regardlessly.

You'd only have to set a few parameters:

- check interval (as is now)
- dynamic yes/no
- maximum interval (after which the next check will be forced)

Then there might be the case that some SMART parameters always change between checks, or that the temperature always toggles between two values (e.g. 21/22 degrees). For this you could either exclude specific values from the "postpone-check-evaluation", or define a "no-difference range" - meaning, a value will only be considered "changed" if it is different by a certain delta.

How does this sound to you?

Regards,
VGER

[hdsentinel]

Thanks for your thoughts !

To detect the current temperature of the hard disk (on IDE / SATA / USB hard disks but not on SCSI/SAS) it is required to detect the complete S.M.A.R.T. information table. It is not possible to say (from the viewpoint of the software) "I'm interested in temperature only (eg. attrribute #194) but not the others".
Detecting the S.M.A.R.T. information table always resets the idle counter, just like any other disk operation (eg. reading a sector on the disk).

So when the temperature is updated - all other attributes are updated as well.
(some software may allow configuration of different intervals to detect temperature and other values - but in this case they only discard the rest of the detected information (except temperature) causing unneccessary high overhead).

I like the idea of the timetable you mentioned: to check status more often on "active" periods and less often on "idle" periods (eg. night).
The dynamic configuration would be fine also, I just think it may be a bit confusing to configure the different x-y-z values and maybe the detection times would not be too clear for the user.

I'm keep thinking about these options

When HDS is started, will it cause sleeping drives to spin up?

Yes. On startup HDSentinel needs to enumerate the drives and check the actual configuration (eg. for changes since last run) and detecting such disk information (not only S.M.A.R.T. but disk position, model, serial numbers and so) wakes up the disks.
This is not a problem if HDS starts with the system, when all drives are still spinning.