Green Caviar: High Load Cycle Cout after short operation time

[hermanboldwick]

I honestly don't get it. Why didn't they allow us to control the aggressiveness via APM or another tool?

I'm using Linux and the LCC increases by almost 100 every hour. What are my options?

* Increase write delay and risk increased data loss in case of power loss.

* Write to the disk every seven seconds to prevent idling at all.

* Use some unsupported wdidle3 tool and void my warranty.

* Use the drive as backup drive and buy something else.

Please tell me, why should I continue to buy and recommend WD drives? 

[christianhberg]

Does that for my WD-20EARS-00MVWB0 also. And i don't even use it for anything. Weird thing the same drive but different revision number (60MVWB0) doesn't even have that problem.

This was the reply i got from WD when i created a ticket:

If both the Caviar Green drives are being used in the exact same way, the load/unload cycle counts should be similar for both. However, if both drives are in the same computer it's extremely unlikely that they're being used the exact same way, This is because you must have different applications running on the PC which require either of the two drives causing the load to be different on both. Please let us know how these drives are configured in your PC; are you using a RAID controller, do you have 1 or more operating systems installed, is one used for data and the other for the OS?

Also, the load/unload cycle count is really not more than just a number. The Caviar Green drives are designed to park the read head as often as they can to save on power consumption. They should work properly even when the cycle count reaches a million or higher. Apart from the increased cycle count, do you notice any performance difference in the drives? If not there is nothing to worry about as the drives are working as expected.

Both drives are being used just as much but the 60MVWB0 have a Load Cycle Count of just 79 and 642 Power On hours but the 00MVWB0 have far over 30.000 thousand with a Power On of 3061 hours. It doesn't take a idiot to notice something is very wrong here.

No background services or other tasks is being used without my knowledge and the work distribution is exactly even between the two.

[gradinaruvasile]

---

hermanboldwick wrote:

I honestly don't get it. Why didn't they allow us to control the aggressiveness via APM or another tool?

I'm using Linux and the LCC increases by almost 100 every hour. What are my options?

* Increase write delay and risk increased data loss in case of power loss.

* Write to the disk every seven seconds to prevent idling at all.

* Use some unsupported wdidle3 tool and void my warranty.

* Use the drive as backup drive and buy something else.

Please tell me, why should I continue to buy and recommend WD drives? 

---

You can use my script:

http://community.wdc.com/t5/Desktop/Green-Caviar-High-Load-Cycle-Cout-after-short-operation-time/m-p...

- essentially you have to poll the drive for SMART data every 10 seconds. This way i dont have any increase in LLC since i use this script.

If you use graphical environment, make sure you do something about  your ~/.xsession-errors file because the output of every smart command will be written into that.

A simple workaround is to delete that file and redirect it to /dev/null:

ln -sf /dev/null $HOME/.xsession-errors

And youre done with it. 

Now i have 225 DAYS of power on hours (thats ~5400 hours) and LLC of 11578. No increase since a few months. The high (ish) number is because at first i wasnt aware of this issue and it grew fast until i made this script.

[christianhberg]

42203 LCC and still counting.

[Tomas73]

Because of high load/uload cycle on my WD20EARX,

I have changed timer with wdidle3 and made a mistake because I changed timer on all of three drives in my rig.

I also have WD500AAKS and WD640AALS.

Timer is now set on 300sec.(5min.) on all my drives and I wanted to do so just on WD20EARX.

Can anyone tell me what is default load/unload cycle time for WD500AAKS and WD640AALS, I want to roll back

timer on default for those drives.

[fzabkar]

Examine the raw values of the SMART Power On Hours and Load/Unload Cycle Count attributes for each drive. Then calculate the autoparking frequency from that.

For example, if a drive has racked up 1000 Power On Hours with 20,000 Load/Unload Cycles, then it is autoparking every 3 minutes on average.

[Tomas73]

They have pretty low numbers of parking, about 0,17 per hour when I divide load/unload cycle number with

operatinal hours.

The default could be "disabled" and that 0.17 per hour could be only when the drive is spinning up or something,

I dont know.

I hope someone have WD500AAKS and/or WD640AALS and help me to find
default values for those drives.
It is simple to check them, just type "wdidle3/R" and the information will show up.
I am trying to find those info. for a week, reading different forums but no luck.

[christianhberg]

Im sorry but how do you get it to 3 min? I don't understand how you can count the autoparking value from Power On Hours and Load/Unload Cycles. Also how do you get RAW values to Decimal values?

[DavidSucesso]

i can see my wd5000AAKS SE 16 later

[Tomas73]

Thanks, I will check the thread later!