Experiment

[Franklin]

Given the conflicting claims, I'm running a current experiment with WD10EACS (Firmware 01.01B01) drives on a DS508 to see if I can make the LLC count go "abnormally" high per hour.

It will consist of creating a RAID-5 volume, and performing an infinite loop of dumping pictures onto the system (dumping pictures creates a lot of HDD activity, given indexing and all).

Alternatively, I may consider torrenting a lot of as well, to see if that changes the behavior.

Also, I'm running the same experiment with a regular computer with a single drive to see if the LLC count can go up

Users should reply with the following if experiencing a high LLC per hour count

1. Screen shot of the Smart Report
2. Synology system System Model Number
3. Synology Firmware
4. HDD Model Number for each drive
5. HDD Firmware for each drive
6. HDD POH for each drive
7. HDD LLC for each drive
8. Services running on Synology system
9. 3rd party software running on Synology system if applicable
10. Ports forwarded to WAN on Synology system
11. RAID Type per HDD


xx. More parameters to come, if needed

[tw_switzerland]

Hi Franklin

First of all: thank you for your initiative on this topic! I'm glad to see that.

Of course I will post here the requested results for all my 4 WD disks as soon as possible.

In your experiment you want to stress the NAS with heavy activity load. But the problem of the dramatic growth of unload/load cycles primarily occurs when the NAS is idle. So you should test that idle case with priority.
The high LC counts seem to be less a problem in systems that are under load. I hope that you have a WD10EACS which shows the problem. It seems that there are versions of the disk that are not affected.

Thanks again.

Thomas

[fabitux]

Hi everybody !
Sorry for my english (I'm french and not good at english speaking)

1. Screen shot of the Smart Report
sda


sdb
By fabitux at 2008-11-28
2. Synology system System Model Number
207+
3. Synology Firmware
DSM 2.0-0722
4. HDD Model Number for each drive
Two WD10EACS-00ZJB0 version 01.0 (1 To each)
5. HDD Firmware for each drive
01.01B01
6. HDD POH for each drive
158 - 158
7. HDD LLC for each drive
now : 0 (zero) but the LCC-problem isn't solved (I will explain later)
8. Services running on Synology system
If I'm running bitorrent : zero LCC on each drive : 0 between a lot of time (and I'm afraid that it isn't a good thing that there are any LCC more ; I would prefer a result like 2,3 LCC per hour for example !!)
9. 3rd party software running on Synology system if applicable
Yep, I add the src synoware http://ipkg.nslu2-linux.org/feeds/optwa ... s/unstable in vi /opt/etc/ipkg.conf. And it seems it works because I have installed 'hdparm' (but hdparm doesn't work on the syno like on my linux laptop... grrrrr !!)
10. Ports forwarded to WAN on Synology system
hmmm I'm not specialist on network... but is this really important for LCC problem ?
11. RAID Type per HDD
Now I am on JBOD. But if I can not solved the LCC problem, I will (try to) migrate on a RAID 1 solution

The end of my post, and "why I can have zero LCC" is posted on http://www.synology.com/enu/forum/viewtopic.php?f=124&t=11653
à+ / fabitux

[tw_switzerland]

Hello Franklin

Here are my values I currently have. I will complete this data later on, as soon as available.

1. Screen shot of the Smart Report
HD1: HD2: HD3:

2. Synology system System Model Number
DS-209+

3. Synology Firmware
DSM 2.0-0728

4. HDD Model Number for each drive
HD1: WD10EACS-00C7B0 (internal 1, volume1)
HD2: WD10EACS-00C7B0 (internal 2, volume2)
HD3: WD10EACS-00D6B1 (external esata, no volume)
HD4: Later

5. HDD Firmware for each drive
HD1: 01.01B01
HD2: 01.01B01
HD3: 01.01A01
HD4: Later

6. HDD POH for each drive
HD1: 282
HD2: 282
HD3: 35
HD4: Later

7. HDD LLC for each drive
HD1: 31'366 (start_stop_count 9) --> LCC problem
HD2: 31'831 (start_stop_count 6) --> LCC problem
HD3: 12 (start_stop_count 12) --> see comment below
HD4: Later

8. Services running on Synology system
Ez-Internet - DDNS
Terminal - SSHD

9. 3rd party software running on Synology system if applicable
None

10. Ports forwarded to WAN on Synology system
None

11. RAID Type per HDD
HD1-4: All basic non raid

COMMENTS ON HD3:

HD3 is a different version of the WD10EACS (called WD10EACS-00D6B1): It has a different case, different labels, different version number, different firmware. And it doesn't show the high load cacle counts. In fact LCC seems to be exactly identical with the start stop counter... (Gee! Did they simply "eliminate" the load cycle counter and show the value of the start stop counter instead?)

First I thought that the external case and the connection via eSata would be responsible for the better LCC behavior of the WD10EACS-00D6B1. But this morning I swapped it with one of the internal WD10EACS-00C7B0. Result: Both disks behave like before. The WD10EACS-00C7B0 now in the external case still suffers from fast increasing LCC (+2 per minute) and the internal WD10EACS-00D6B1 does not increase LCC at all.

The following two questions are in the room now:
Is the newer WD10EACS-00D6B1 version improved and the LCC problem is really absent?
Or does the new firmware simply show another vlaue (start stop counter) as load cycle counts?

Many thanks. I hope we can collect enough data to identify the problem.

Thomas

[Volpone2]

Given the conflicting claims, I'm running a current experiment with WD10EACS (Firmware 01.01B01) drives on a DS508 to see if I can make the LLC count go "abnormally" high per hour.

....

Great idea Franklin ! Bravo !

Some precisions :

1) High load LCC issue is "opened" since march 2008, IMHO all Synology models and WD Green Power (500.750, 1000 Go models) are concerned (all share "Intellipark" technology).
2) Some users reports that recently released WD GP models with a new firmware reference fix the problem.
3) This is not a specific Synology issue but a Linux / Intellipark issue.
4) Users are waiting information and recommandation from Synology and not only to list Western Digital Green Power disks as compatible.

Regards

[Volpone2]

....
COMMENTS ON HD3:

Disk HD3 is in an external box and connected to the front esata-connector of the DiskStation.
It is remarkable that HD3 at first sight seems not be affected by the high LCC problem. As this disk is another version
with other firmware and attached differently, I will have to swap it with one of the internal harddisks.
Otherwise it will not be possible to say if the difference is due to the external connection or the other firmware.

Thomas
IIve mounted a WD GO 750 Go drive in an external E-sata enclosure => No more high LCC issue
Same drive mounted internally as drive 1 or drive 2 => High LCC values (same as yours)

So i'm curious about your measurement with your new WD GP firmware mounted internally.

Regards

[thermal]

This is not meant to confuse the matter, but here's one for the WD1000FYPS-01ZKB disks, the enterprise grade HDD's, that seem to have the same symptoms:

1. Screenshots: <can't add attachments in this forum>
2. DS207+
3. Firmware 2.0.0722
4. WD1000FYPS-01ZKB (2x)
5. Disk's firmware 02.01B01 (both)
6. POH disk1: 1369. disk2: 1369
7. LLC disk1: 14.760. disk2: 14.673
8. Services: Only Telnet and WIN OS, the device is for Backups only.
9. 3rd party software: None.
10. Ports forwarded: None
11. Raid 1

As said, this DS207+ is used for backup purposes only and is therefore idle most of the time.
The disks do hibernate, which is set to 10 minutes.

[tw_switzerland]

Thank you thermal for your data.

Remark: If your system runs idle most of the time, then your system seems not to be affected by the problem. An increase of the LCC of +10 per hour is just fine and should not be a problem.

See instruction on how to upload images. (Remark: You must have at least 3 posts in order to be allowed to link in images.)

[dizzzz]

update; after some observation and calculation I see an LCC increase of 80+ per drive per hour. this means that my drive will die in 160 days...

1. Screen shot of the Smart Report
(not possible yet)

2. Synology system System Model Number
209+

3. Synology Firmware
DSM 2.0-0728

4. HDD Model Number for each drive
a&b: WD10EACS-00C7B0

5. HDD Firmware for each drive
a&b: 01.01B01


6. HDD POH for each drive
a: 72hr
b: 53hr

7. HDD LLC for each drive
a: 988 in 72 hrs
b: 333 in 53 hrs (drive is one week newer)
and counting fast.....

8. Services running on Synology system
PhotoSystem, iTunes server

9. 3rd party software running on Synology system if applicable
None

10. Ports forwarded to WAN on Synology system
None

11. RAID Type per HDD
Raid1

[dizzzz]

+40 for each drive when I typed the message

[GNaschenweng]

Running on DS107+ and latest firmware. Got iPKG installed and a number of cronjobs. The disk never goes into hibernation and my LCC increased from yesterday 12:51pm (319814 LCC) to today 07:40am (323208 LCC) - thats 170 LCC/per hour:

Code: Select all

=== START OF INFORMATION SECTION ===
Device Model:     WDC WD10EACS-00ZJB0
Serial Number:    WD-WCASJ1111774
Firmware Version: 01.01B01
User Capacity:    1,000,204,886,016 bytes
Device is:        Not in smartctl database [for details use: -P showall]
ATA Version is:   8
ATA Standard is:  Exact ATA specification draft version not indicated
Local Time is:    Sat Nov 29 07:30:29 2008 CAT
SMART support is: Available - device has SMART capability.
SMART support is: Enabled

=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

SMART Attributes Data Structure revision number: 16
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   200   200   051    Pre-fail  Always       -       0
  3 Spin_Up_Time            0x0003   185   183   021    Pre-fail  Always       -       7750
  4 Start_Stop_Count        0x0032   100   100   000    Old_age   Always       -       692
  5 Reallocated_Sector_Ct   0x0033   200   200   140    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x000e   200   200   051    Old_age   Always       -       0
  9 Power_On_Hours          0x0032   096   096   000    Old_age   Always       -       3219
 10 Spin_Retry_Count        0x0012   100   100   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       -       59
192 Power-Off_Retract_Count 0x0032   200   200   000    Old_age   Always       -       52
193 Load_Cycle_Count        0x0032   093   093   000    Old_age   Always       -       323208
194 Temperature_Celsius     0x0022   110   104   000    Old_age   Always       -       42
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


[botho]

Hy,

First of all, many thanks for investigations.

1. Smart Report :
HDD1


HDD2


2. System Model Number : DS207+

3. Synology Firmware : DSM 2.0-0728

4. HDD Model Number :
HDD1 & HDD2 : WD10EACS-00ZJB0

6. HDD POH :
HDD1 : 2391
HDD2 : 2391

7. HDD LCC :
HDD1 : 97185
HDD2 : 93166

8. Services running :
AppleTalk (afpd)
Photo Station / File Station / Audio Station
SSH
Download Station (only rtorrent)

9. 3rd party software running :
IPKG (KISSDX deamon for video streaming to LINKSYS KISS1600)

10. Ports forwarded to WAN on Synology : none
Is it to WAN or from WAN ?
if Syno to WAN : rtorrent TCP and UDP forwards only

11. RAID Type :
HDD1 & HDD2 : Basic volumes

[iolo]

Good to see this problem taken into account...

1. Smart Report :
HDD1:
HDD2:

2. System Model Number : DS207+

3. Synology Firmware : DSM 2.0-0728

4. HDD Model Number :
HDD1 : WD10EACS-00C7B0
HDD2 : WD10EACS-00D6B0

5. HDD Firmware
HDD1: 01.01B01
HDD2: 01.01A01

6. HDD POH :
HDD1 : 3596
HDD2 : 1715

7. HDD LCC :
HDD1 : 12053
HDD2 : 224
I have used the wdidle on the HDD1 to raise the timeout to the maximum (25s)
No problem with the second drive (more recent), I have never used Wddidle on it.

8. Services running :
SMB, NFS, SSH, UPNP
File Station / Audio Station / Download Station

9. 3rd party software running :
IPKG with Firefly (efficient way to stream audio to my Roku M1001, by the way it crashes if synology indexing service is running).
All IPKG stuff is on a USB key following this tutorial

10. Ports forwarded to WAN on Synology : none
SSH and torrent

11. RAID Type :
HDD1 & HDD2 in RAID 1

I should add that I have applied with very little effect all tricks around here for my faulty HDD1 (noatime, expicentisec, USB key, etc)
All my LLC in this drive are the result of the SMB accessing every minute or so the drive, even when nothing is happening on the connected PC and MAC.
If Synology manages to have this regular SMB access to not access the drive (with some cache I don't know), it should solve the issue, at least in my case. For me, this drive is just incompatible with a SMB network access (same kind of trouble with AFP by the way). NFS behaves correctly in this area (but unfortunately, it does not answer all my needs).

Hibernation works fine for me.

Hope this help.

[tw_switzerland]

Hi iolo

It seems like beginning with WD10EACS-00D6B0 version of the drive, the high LCC are gone. Question is, if they simply show another counter as LCC (i suppose start stop counter) now.

I don't think SMB is the problem here. As I have +120LCC per minute, I suppose the sync process to be a potential problem. If I am right disks are synced automatically every 30 seconds in a standard linux system. But I am not 100% sure on that.

Thomas

[iolo]

It is a very good point indeed. I did not notice that before but as you mention LLC on newer drive is now equal to the SSC!!!!
Very strange! Either they have hidden the problem (that is still here), either the feature is not running anymore (this obviously solves the problem).
In both case, great news
Thanks to point this out.

Regarding the linux sync process, you can reduce this 30s to whatever you want.
Just put (for example) "/sbin/sysctl -w vm.dirty_expire_centisecs=700" in some boot script to have it at 7s instead of 30s.
After you have done that, you will see (with help of the "syno_hibernate_debug_tool --enable 10" command) that SMB accessing regurlarly the drive, for nothing, preventing hibernation when other computer are running on the network and doing a LOT of LLC.