Seagate Ironwolf noise and increasing Load_Cycle_Count

Looking for discussion about hardware compatibility? Look no further.

Moderator: Lillian.W@AST

foggy69
Posts: 20
youtube meble na wymiar Warszawa
Joined: Thu Mar 09, 2017 7:46 pm

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by foggy69 »

vitosx wrote:T
According to information on the web Seagate issued a firmware update SC61 fixing the issue for many users of other NAS devices, but only for 6TB and 8TB units. Just out of curiosity I would ask you to check if this is a firmware version you have on your drives, but it is not easily done from ADM web interface. You might need to use commandline smartctl or hdparm tool for that. If this is a trouble, don't bother. I guess it would not help me much either.
Checked my ST8000VN0022-2EL112 discs, and they are both running SC61 firmware.
vitosx
Posts: 52
Joined: Sun Sep 24, 2017 11:30 pm

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by vitosx »

foggy69 wrote:Checked my ST8000VN0022-2EL112 discs, and they are both running SC61 firmware.
Thank you.
So something is up, I wonder why Seagate haven't addressed this issue on smaller drives yet. Maybe they don't get enough complaints from customers.
sksbir
Posts: 395
Joined: Tue Aug 25, 2015 9:23 pm

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by sksbir »

smartctl /dev/sda -a|grep -i firm
Firmware Version: SC61

on my both NAS.
honcho
Posts: 1
Joined: Wed Nov 22, 2017 12:01 am

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by honcho »

Any update to this thread, I'm having the same problem with 2 WD green drives.


NAS is idle, I have no apps installed, and I can hear one or both of the drives do a random seek, then park the heads. Over and over and over and over. Logs indicate drives are in hibernate, but they most certainly are not.

One of these drives has been in an external enclosure attached to a WD Mycloud via USB for over a year.. and it's Load_Cycle_Count tripled in 2 weeks being in this stupid thing. For the time being I have it turned off and I just turn it on when I watch a movie.
vitosx
Posts: 52
Joined: Sun Sep 24, 2017 11:30 pm

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by vitosx »

No real update.

Here's what Seagate support had to say:
You are not the first to contact us about this activity in the IronWolf drive. It does seem to worry owners who are not used to this. The purpose of the frequent starts and stops is that the built in firmware is doing exactly what you said it sounds like. It is doing a mini defrag every time it does this.

The purpose is so that it can intelligently arrange the data on the disks in such a way that it is tightly packed with little to no wasted space and with the files compacted in such a way that they are very fast to both find and transfer to where ever they are being called to. It does make the unit a little noisy but it is not damaging to the drive. This is part of what is referred to in our webpage concerning the Health Management of the drive. These disks are specially built to take the wear and frequent usage and are meant to run 24/7.
I won't comment on that. I just made my point that too frequent head parkings can hardly be considered "healthy" and they promised to forward my feedback to proper "experts".

Asustor support did not comment nor replied to this case.

For now I don't recommend this specific 4TB model of IronWolf to be used in Asustor NAS.
I haven't replaced it though, as at least I found satisfactory workaround, but I will post the details when I have access to my NAS. I borked my VPN access somehow...
MonsMagnus
Posts: 74
Joined: Sat Mar 11, 2017 2:17 am

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by MonsMagnus »

vitosx wrote:No real update.

Here's what Seagate support had to say:
You are not the first to contact us about this activity in the IronWolf drive. It does seem to worry owners who are not used to this. The purpose of the frequent starts and stops is that the built in firmware is doing exactly what you said it sounds like. It is doing a mini defrag every time it does this.

The purpose is so that it can intelligently arrange the data on the disks in such a way that it is tightly packed with little to no wasted space and with the files compacted in such a way that they are very fast to both find and transfer to where ever they are being called to. It does make the unit a little noisy but it is not damaging to the drive. This is part of what is referred to in our webpage concerning the Health Management of the drive. These disks are specially built to take the wear and frequent usage and are meant to run 24/7.
I won't comment on that. I just made my point that too frequent head parkings can hardly be considered "healthy" and they promised to forward my feedback to proper "experts".

Asustor support did not comment nor replied to this case.

For now I don't recommend this specific 4TB model of IronWolf to be used in Asustor NAS.
I haven't replaced it though, as at least I found satisfactory workaround, but I will post the details when I have access to my NAS. I borked my VPN access somehow...

I would be very interested in seeing your work around, please post when you get a minute.
vitosx
Posts: 52
Joined: Sun Sep 24, 2017 11:30 pm

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by vitosx »

My IronWolf stopped its nasty behavior after I changed Advanced Power Management (APM) level to maximum allowed 254 value (meaning: disable auto power saving) using hdparm commandline tool. Unfortunately this change makes the whole NAS to hang (with HDD LED blinking) if it tries to enter S3 standby mode, or on shutdown/reboot. So it needs to be reverted to default 128 value before shutdown. I could not overcome standby issue, so Sleep Mode must be disabled in Energy Control settings. HDD fortunately still goes into Hibernation as configured.

Here are the steps.
  • you need to install entware-ng package from App Central, if you don't have it already, in order to get hdparm
  • login to NAS using ssh as root (the password is the same as for admin user, SSH service needs to be enabled of course)
  • install hdparm by running

    Code: Select all

    opkg update
    opkg install hdparm
  • test if it actually changes anything in head parking behavior for your drive by running

    Code: Select all

    hdparm -B254 /dev/sda
  • if it does, then create scripts in /usr/local/etc/init.d directory to automate things
  • launched on startup:

    Code: Select all

    echo "#!/bin/sh
    sleep 5
    /opt/sbin/hdparm -B254 /dev/sda" > /usr/local/etc/init.d/S99disable-apm
    chmod +x /usr/local/etc/init.d/S99disable-apm
  • and on shutdown:

    Code: Select all

    echo "#!/bin/sh
    /opt/sbin/hdparm -B128 /dev/sda
    sleep 5" > /usr/local/etc/init.d/K99default-apm
    chmod +x /usr/local/etc/init.d/K99default-apm
Notes:
  • /dev/sda is device path; I have only one drive at the moment in my NAS so it is easy, if you have more devices you must find the proper one and I don't know if drives are getting the same identificators on every reboot
  • sleep 5 command slowing down the script execution may not be necessary but I had inconsistent behavior without it
  • not every HDD allows to change APM level and not every reacts the same for specific values
MonsMagnus
Posts: 74
Joined: Sat Mar 11, 2017 2:17 am

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by MonsMagnus »

vitosx,
Thank you very much for the detailed reply! I will test this as soon as possible.

Regards
sksbir
Posts: 395
Joined: Tue Aug 25, 2015 9:23 pm

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by sksbir »

vitosx
a nice reverse engineering action ! 8-)

Once your APM is set to 254, is the disk always running even if led is blinking ( that is : disk ignores power saving requests from NAS ), or will it still go to hibernation if NAS requests to ?
vitosx
Posts: 52
Joined: Sun Sep 24, 2017 11:30 pm

Re: Seagate Ironwolf noise and increasing Load_Cycle_Count

Post by vitosx »

Yes, it will go to hibernation. The disk just does not park its heads any more and what is as much important - does not do its magical noisy heads seeking activity. Which is a little irony, because according to my energy meter, NAS uses a little less energy now. It draws 0,5 to 1W more when heads are actively seeking, and I did not measure any energy savings when heads were parked vs normal idle. Hibernation lowers NAS energy usage by 4W, so it is a reasonable benefit.

Actually I think NAS does not issue any request, but by setting Hibernation time in ADM setttings you are setting internal idle timer in the drive itself (you might get same results by using hdparm -S command). That's the reason some USB attached HDDs (like WD Desktop Elements) might not hibernate at all when connected to Asustor NAS (they do not support this command, but they work with a tool like hd-idle, which is unfortunately not available in entware-ng nor App Central.
Post Reply

Return to “Hardware Compatibility”