WD40EFRX - LoadCycleCount issue is back?!

[mnuesser]

ok, got it to work by my self

this is how i have done it without pc (sorry for my bad english)

1. i setup my system from ground on, so first insert just the hdd1, do all my basic setup like network...
2. open the web-frontend, open the app-center, install "Optware IPKG"
3. open ssh, connect to qnap, then "PATH=$PATH:/opt/bin" <- makes it easier, is not permanent
4. still ssh: "ipkg update", "ipkg install mc" <- i use this a lot...
5. still ssh: "ipkg install gcc", "ipkg install automake"
6. still ssh: "wget http://sourceforge.net/projects/idle3-t ... -0.9.1.tgz"
7. still ssh: "tar xzvf idle3-tools-0.9.1.tgz", "cd idle3-tools-0.9.1"
8. still ssh: open mc, select the Makefile in folder, press F4 for edit, go to line 13 and change "CC ?= gcc" to "CC = gcc", save and exit mc
9. still ssh: "make -F ./Makefile" - now you have the compiled version, try to start to see if its working "./idle3ctl -V" it should say something like V 0.9.1
10. to keep this file you can open mc again, and copy that idle3ctl file to /opt/bin...
11. shutdown the qnap
12. insert new disks
13. startup qnap, wait until completely booted up, connect via ssh again
14. i tested now if the program is working with reading actual value of the idle timer from my drive hdd2 (/dev/sdb)
"/opt/bin/idle3ctl -g105 /dev/sdb" and the programm tells me that the setting is 8 sec.
15. change the value to disable for that drive: "/opt/bin/idle3ctl -d /dev/sdb", to get this new value active, you have
to shutdown the system... the hdds must be turned off completely to set the new timer value...
16. after starting up again, you can check the new value with "/opt/bin/idle3ctl -g105 /dev/sdb" again, it should tell
you: "Idle3 timer is disabled"

thats it...
as you see, i am not very good in compiling, but maybe someone can tell me, if i can give my compiled file to someone else,
with the same kind of cpu (INTEL), so they just have to start that file, instead of compile it by there own...

here you find some more details to that small software tool:
http://idle3-tools.sourceforge.net/

[spindoctor]

mnuesser - thanks a ton for your handy tip to do this on the Qnap! - Works fine for me and saved me from shuffling it to/from a PC

To confirm -

I have a TS-569L and received 5x WDC WD40ERFX-68WTON0 drives, all at WC firmware version 80.00A80 today and noticed LCC increasing at quite a rate as I did back in the day with WD Green drives.

I largely followed mnuesser's instructions with a few I skipped as I'm a bit more comfortable at the command line (ie. if you are not comfortable follow mnuesser's steps):
* I did not install 'mc' as I'm quite comfortable with 'vi' which is pre-installed and it is only needed to edit the Makefile
* I had all my drives inserted at the time
* I retrieved the idle3-tools tarball, edited the Makefile, compiled it and did not copy the resulting binary in /opt/bin/ as I only wanted to use it once on the system (seems like /root gets blanked on reboot)
* All my drives had a value of 80 by default ):
** Just as a FYI Drive1 = /dev/sda, Drive2 = /dev/sdb, Drive3 = /dev/sdc, Drive4 = /dev/sdd, Drive5 = /dev/sde, you obviously need to do this with all your WD40ERFX drives
** As was mentioned - everything for this binary is self-explanatory at http://idle3-tools.sourceforge.net/
* idle3-tools tells you to shutdown and restart (not reboot) for changes to take effect, on my qnap model a reboot did the trick, so I assume the reboot process on qnap power downs the drives at some point...

I have put the compiled binary at https://www.dropbox.com/sh/0etjaxu4u2l9msr/FBkgeewsSE for those who do not want to go through the compiling drama.
* Copy the file to the Qnap, mark it as executable (chmod +x <filename>) and run it...

[dan2507]

ok, got it to work by my self

...

Thanks a lot for the good instruction, was easy to follow and now I can fix the issue directly on the NAS without the need to move this to PC 1st.

[dan2507]

I can confirm that the current delivered WD40EFRX have the idle timer set to 300 sec.

S/N: WD-WCC4E075****
FW: 80.00A80

I have still deactivated it, I don't see any reason to have a value there.

[fabiocasi]

I have 3 WD40EFRX production 09,13,18 september 2013 and one WD10EFRX 12 december 2013. ALL set to 8 second!!!

I would like to understand if is BETTER to disable WDIDLE3 /D or set to to 5minute (300 seconds) WDIDLE3 /S300 the disks for a NAS.

In the past, with GREEN series, the /d option wasn't suggested because can generate problem on the disk!

[schumaku]

I have 3 WD40EFRX production 09,13,18 september 2013 and one WD10EFRX 12 december 2013. ALL set to 8 second!!!

Yes, still the same default.

I would like to understand if is BETTER to disable WDIDLE3 /D or set to to 5minute (300 seconds) WDIDLE3 /S300 the disks for a NAS.

Ask WDC - wrong place here.

In the past, with GREEN series, the /d option wasn't suggested because can generate problem on the disk!

...if the NAS drops from the rack, or somebody kicking it - the heads are parked after eight seconds.

On both Green and Red - I've never seen an answer if a high LCC is an issue at all. Certainly, there will be more head carrier movements, some more noise, potentially some lower performance, and a counter changing.

Again - please talk to WDC, and let us know if you go the perfect information.

[dan2507]

I have 3 WD40EFRX production 09,13,18 september 2013 and one WD10EFRX 12 december 2013. ALL set to 8 second!!!

Yes, still the same default.

Not true, the drive I've bought yesterday has a timer of 300sec - did not check production date, but pasted the S/N info above.

I'd also be interested what the advantages/disadvantages are for 300 sec compared to disabled. As the Red 2 + 3TB drives has this timer disabled it sounds logical to keep this, I hope I'm correct.

Another thing, after I've used the cmd to disable the timer the program tells me about a needed power off, but reading the timer after changing it I've got the changed settings. So is this warning valid, and if yes is a reboot enough for getting the changes applied?

[schumaku]

Not true, the drive I've bought yesterday has a timer of 300sec - did not check production date, but pasted the S/N info above.

I know QNAP (and probably other NAS vendors) have approached WDC on this unlucky default on the earlier units.

I'd also be interested what the advantages/disadvantages are for 300 sec compared to disabled. As the Red 2 + 3TB drives has this timer disabled it sounds logical to keep this, I hope I'm correct.

General information is that the HDD will unload (heads parking or the like) after an idle time of 300 s now. Depending on your NAS usage, the LCC count will still grow (just slower than with 8 s) If disabled, the heads are never parked while run time, only when shutdown or unexpectedly powered off. Again, challenge WDC if you want to know more.

Another thing, after I've used the cmd to disable the timer the program tells me about a needed power off, but reading the timer after changing it I've got the changed settings. So is this warning valid, and if yes is a reboot enough for getting the changes applied?

Changing and reading does access the permanent storage, not the volatile or runtime values. A reboot does neither power-off nor temporary cut the NAS HDD power - the controller remains booted with it's previous configuration values. That's why they ask for a power-off power-on sequence.

[dan2507]

I have 3 WD40EFRX production 09,13,18 september 2013 and one WD10EFRX 12 december 2013. ALL set to 8 second!!!

Yes, still the same default.

... did not check production date....

Just checked it, its 12DEC13, so at least this production date has the idle not anymore at 8 but 300 sec., at least for the WD40EFRX.


@schumaku: Many thanks for the explanations

[scorpydude]

What should the idle time be on a "nas" grade drive ?

[ihartley]

Not 7 seconds Low enough for it not to affect performance, low enough to provide power saving based on your usage/needs, high enough that the LCC doesn't reach the stated specification over its life.

Personally - unless you're running gazillions of them - I'd set to 300s or never. Newer drives are operating at sub-5 watt, so cost saving on power in setting a lower time is negligible. I'd imagine - though I have no data to back it up - the power saving would be offset by the increased failure rate of drives that spin up often (where there is most stress on the motor), combined with the cost of replacement of failed drives/data as a result.

[schumaku]

...s, high enough that the LCC doesn't reach the stated specification over its life.

And what specifications are you talking of here?

PS. Can't get rid of the impression some 2 TB Red (NASware 2.0 - WDC WD20EFRX-68EUZN0) added to some NAS tend to start to count up there, too.

[dan2507]

For the RED series the max LCC count is 600.000, for the Green 300.00o if I remember correct.

[schumaku]

For the RED series the max LCC count is 600.000, for the Green 300.00o if I remember correct.

References? WDC spec sheets?

[Scillonian]

For the RED series the max LCC count is 600.000, for the Green 300.00o if I remember correct.

References? WDC spec sheets?

From an August 2012 spec sheet (PDF) I have for WD Red drives the Load/Unload Cycles is quoted as 600,000.

Update: The August 2013 version also quotes the same value.