Load / Unload Cycle_Count increases quickly on WD GP Caviar

[Volpone2]

Hi,
I'm using 2 WD 7500AACS drives (Western Digital Green Power Caviar) on DS207+, RAID 1 , FW 518.
All seems correct, temp is 32C to 37C and not much noise.

I'm monitoring the drives via "Moniteur-Web" script using Smartctl ...
I'm worried about Load_Cycle_Count parameter which increases very quickly:
203394 Load - Unload cycles / 267 Power on hours : ie 1 cycle every 40" !!!
Western Digital specs for this parameter is 300 000 max ! So my drives are near to be "old" !

WD IntelliPark is perhaps too efficient and agressive in "unloading the heads during idle to reduce aerodynamic drag" or there is some issue with synoware ?
I' ve read some very interesting thread about this in Ubuntu forum, perhaps the same problem with Syno OS ?

Questions:
- Anybody with same issue ? (there is post in french syno nas-forum)
- Risks for the drives ?
- How to fix it ?

Smart parameters (excerpt)
....
7 Seek_Error_Rate 0x000e 200 200 051 Old_age Always - 0
9 Power_On_Hours 0x0032 100 100 000 Old_age Always - 267
10 Spin_Retry_Count 0x0012 100 253 051 Old_age Always - 0
11 Calibration_Retry_Count 0x0012 100 253 051 Old_age Always - 0
12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 9
192 Power-Off_Retract_Count 0x0032 200 200 000 Old_age Always - 7
[b]193 Load_Cycle_Count 0x0032 194 194 000 Old_age Always - 20394[/b]
194 Temperature_Celsius 0x0022 116 100 000 Old_age Always - 36
196 Reallocated_Event_Count 0x0032 200 200 000 Old_age Always - 0
197 Current_Pending_Sector 0x0012 200 200 000 Old_age Always - 0
198 Offline_Uncorrectable 0x0010 200 200 000 Old_age Offline - 0
199 UDMA_CRC_Error_Count 0x003e 200 200 000 Old_age Always - 0
200 Multi_Zone_Error_Rate 0x0008 200 200 051 Old_age Offline - 0
...

Regards.

Volpone

[Franklin]

Greetings Volpone2

Our hardware engineer looked at this thread and found the following math....

Load Count = 20394 (Not 203394)
If 20394 / (Power On Hours) = 20394 / 267 ==> 76 Per Hour ==> 1.27 Per Minute ==> Normal? Abnormal? Up for interpretation



Though we are looking at the Ubuntu Wiki for more information, however our system uses a Linux Kernel, (not any flavor of Linux) so it may not apply to our systems, but we are looking into your question.

Is this number constantly changing every day?

[Volpone2]

Hi Franklin,
I agree with the math ... 1,27 Load / Unload cycle per minute.
MTBF (see WD specs for this disk) is max 300 000 Load / Unload Cycles, so in 4 month my 2 Caviar GP have a risk to be dead !

SMART reported parameter is increasing on the same high basis every day, status today:
9=> Power on hours: 294
193 => Load_Unload_Cycle: 22501

That's a lot unecessary park / unpark operations. I've tested without any Diskstation activities (disconnecting all clients) and nothing is different.
Apart from Ubuntu Wiki there is a lot of ressources about this issue on the net (search Google).
All seems about Linux OS associated with "agressive power management new generations drives" (WD Intellipark and Hitachi own techno).
No problems reported with windows attached drives.

Let me know
Regards

[Cinhil]

Hi,

It is the same behavior with the WD10EACS GP 1To... I had 1,45 cycles per minute.
I have stoped the increase by using an other classical disk as volume1 and the WD as volume2 but it avoid any use of Raid arrays.
There is no difference for this problem between the 518 and the 571...

Regards,
Cinhil

[Volpone2]

I just posted a step by step "how to" about Smartmontools installation on a DS207+ / FW518 in the french Synology forum (sorry but only in french at this time)
=> http://www.nas-forum.com/forum/t3947.html

I will be usefull for the Syno community if WD Green Power disks users use it to check if very high "Load / Unload Cycles Count" is a general issue with WD new generation Green Power disks.

Regards

[Franklin]

Just as a suggestion/Discussion, it is better to judge a a disk health with the amount of defective sectors that the HDD has remapped, or will remap.

If one bad sector is found, the HDD will attempt to remap it, where Reallocated Sector Count and Reallocation Events Count will increase. If the HDD is out of spare sectors for remapping, then the previous mentioned numbers will stop rising, and Uncorrectable Sector Count will begin to increase. This is at the point where the health of the HDD becomes highly questionable.

If a system is remapping a sector in a NON-RAID environment, the data loss may occur on that sector or the entire volume
If a system is remapping a sector in a RAID environment, the disk may be ejected from the RAID Array.

Most SMART utilities usually predict disk life/longevity by the error rate of defective sectors, amount of remaining sectors remaining, and the amount of hours that the disk has been in operation.

[Volpone2]

I agree Franklin about this general information about disk health (Adair from synology support mail me the same elements) but:
- 76 Load / Unload Cycle per hour is a very big and useless stress for the disk mechanic (with no disk activities)
- Disk manufactuer give a maximum of 300 000 cycles specification, in my configuration this is 170 days to reach this theorical limit if disks works 24/24, 7,7...
- A lot of users are worried about this phenomenous that might shorten disk life (see Linux and ubuntu threads about disk killer)
- Hypothesis is that a specific issue to new "power savings" drives like the WD Green Power line

I hope Synology or Western Digital will give a fix or workaround.

Regards

[LosOutlandos]

I hope Synology or Western Digital will give a fix or workaround.

i hope so too! i just bought a western digital WD1000FYPS, which seems to be designed specifically for 24/7 usage, so it would be ironic if such a drive would die more quickly than others because of its power managemant.

[Volpone2]

I hope Synology or Western Digital will give a fix or workaround.

i hope so too! i just bought a western digital WD1000FYPS, which seems to be designed specifically for 24/7 usage, so it would be ironic if such a drive would die more quickly than others because of its power managemant.

LosOutlandos,
I am very interested in Load_Cycle_Count value measurement with your new WD GP drive.
Paradoxically, in my configuration, is when there is no disk activities than this parameters grows very quickly.

Regards

[LosOutlandos]

hi Volpone2!
i'd be glad to do what i can to help figure out what's wrong with the drives.
the drive is still being processed at my vendor, i'm expecting it thursday.
i'll then try to install this "Smartctl" if you could give me some pointers how to do so.
(my french is a little rusty, but i'll try...)

[Volpone2]

hi Volpone2!
... i'll then try to install this "Smartctl" if you could give me some pointers how to do so.
(my french is a little rusty, but i'll try...)

My english is, too, "a little rusty" .
Here is the procedure and pointers i've used to install smartmontools (DS207+ | FW 518)
=> http://www.nas-forum.com/forum/t3947.html (sorry only in french for the moment)

Let me know
Regards

[icy]

I have installed 1 WD 1000FYPS on a DS-207+ and smartmontools counts around 21 load_cycle counts /hour. An second drive is waiting outside the DS-207+ for its install, but under these circumstances it will stay outside

[Volpone2]

Icy,
21 LCC is not a lot, i'm currently more than 76 cycles per hour ! (for 300 Hours Power_ON).
I'm interested if your are monitoring this parameter for some days and see if this value remains the same.

I've submitted a ticket to Synlology an Western Digital about this possible issue.
May be you could also send also a question to both suppliers.

Regards

[icy]

i will call WD U.S support today and put the answer here into the forum.

[LosOutlandos]

i will call WD U.S support today and put the answer here into the forum.

thanks, i`d be very interested in that as i'm also using the 1000FYPS.
i don't know if there's a connection to this issue, but i noticed there were some different jumper setting possible with the drive. maybe this could deactivate the 1000FYPS's power management?