Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[JAylmer]

I own two DNS-343's.  Both in Raid 5.  (4 x WD 1.5TB WD15EARS) and (4 x Samsung 1TB HD103UI).
Over the years I have had three drive failures, the first two were copybook, this one is a drama.

This time, one of the 1.5TB WD drives has failed and dropped out of the array. No data lost just a bit slow.
I looked at the dlink AU web site which shows a drive compatibility list that is > 3 years old.  None of those drive are currently available now.  The supplier could not really help in a choice of new drives but suggested that they have no probs generally with the Seagate Barracuda 1TB ST1000DM003   I took a punt with four of these and what a disaster.  They seemed to format up as a 4 drive raid but when completed there was no data area and two of the four drives had dropped out of the array. I updated the firmware on the DNS-343 to 1.06beta but it made no difference.  I also updated the firmware in each of the drives but still no change.

I ran the SeaTools and ran tests that did not take very long and early terminated those that did.  All tests passed except for the Full Erase which failed immediately on the four drives.
Rang DLink but they could not help as they were not on the compatibility list.  They also said to take the DNS-343 back to firmware 1.04 which is the latest AU supports.  The supplier initially said sorry but they could not accept the drives back. Seagate support is not available during the day, have to ring after 10:00am Singapore time, whatever that is?  
Put a drive in a USB caddy and formatted it up (Win 7) no probs and copied data also no probs.

I connected a drive as drive C: and loaded Windows 2003 on a different machine.  It failed immediately after the format when it tried to copy data across.  Tried another on same result.
I actually purchased 5 of these drives to have a cold spare and it was still sealed.  Reluctantly I broke the sealed bag and tried the Full Erase on this drive which it passed.  Tried it in the Win 2003 install but it also failed.  Worked OK as a data drive in a USB caddy but not as a boot drive and it had not been in the DNS-343.
Found a spare Seagate 1TB  Barracuda ST31000340AS and installed Win 2003 on it no probs and its still going fine.
The supplier agreed that there may be more to it so I have returned them as a warranty claim, but 5 drives failures is remarkable.

I was not sure if the original bad drive was really bad or not so I formatted it (USB caddy) and it came up OK.  Put it back in the DNS-343 and click on regen.  While it was regenerating I could still read the data.  The instant it came on line about 80% of my files dissapeared, those still there were corrupt.  I removed the bad drive and the files were accessible again without any corruption.

The drive bags were sealed with a sticker that said "Made in China" but what isn't.

I now have the 2011 compatibility list but many of the drives listed are old. I am now considering whether to buy 5 x WD Black Caviar 2TB ($1,000).  Any thoughts on replacements?
Are the DNS-343's too old now?

[JavaLawyer]

Here is a compatibility list from March 2011: DNS-343 - HDD Compatibility List.  This list may be predicated on Firmware v1.05+ (I'm not sure). FW 1.05 support advanced format HDDs, but only up to 2TB.

[JavaLawyer]

Regarding the DNS-343 showing its age, yes the product is old, but it still works well (I also own two units).  If you are looking for a new device, the DNS-345 is the successor to the DNS-343: DNS-345. The DNS-345 is a four-bay device that supports 3TB, substantially increasing maximum capacity over the DNS-343.

[JAylmer]

Thanks for the current DNS-343 list.  I will install the 1.05 firmware.

The DNS-345 has a compatibility list that is about the same as the DNS-343.  I don't need > 2TB so I will stay but my concern is that I won't be able to buy compatible drives in the future.

I suspect many people have had success with all kinds of drives but how do we find out?  Even in this forum a few posts have broadly described a drive that was successful but no exact model number or whether is was in a raid or single drive configuration??  Then there is the issue of how it operates when errors occur. If it exhaustively retries then it will prematurely drop out of the raid.  It's a nightmare!

I had a look at the Netgear solutions but same issue with the drive compatibility.

Does anyone know if there are functional differences between the A and B versions of DNS-343 hardware?

[JavaLawyer]

Thanks for the current DNS-343 list.  I will install the 1.05 firmware.

You may want to skip to the latest version 1.06b01, which has a fix to FTP.

The DNS-345 has a compatibility list that is about the same as the DNS-343.  I don't need > 2TB so I will stay but my concern is that I won't be able to buy compatible drives in the future.

I suspect many people have had success with all kinds of drives but how do we find out?  Even in this forum a few posts have broadly described a drive that was successful but no exact model number or whether is was in a raid or single drive configuration??  Then there is the issue of how it operates when errors occur. If it exhaustively retries then it will prematurely drop out of the raid.  It's a nightmare!

The only real way is to read user testimonials for different HDDs.  Personally, all 8 HDDs in my two DNS-343 are in standard configuration, and they have been running for 3.5 years with no errors and no HDD failures.  It's inconvenient to work with smaller volumes, but that is the cost of stability.

Does anyone know if there are functional differences between the A and B versions of DNS-343 hardware?

I'm not 100% sure, but I think the "B" version has a more compact power supply. Not sure if there are internal differences.  Don't quote me on this. 

[JAylmer]

I purchased a third DNS-343 second hand, upgraded the firmware to 1.6b01
Purchased four Western Digital WD2002FAEX Black Caviar drives  2TB  SATA 6G
Installed drives, created a raid 5 array with the four drives.
These drives are on the updated compatibility list.  All working fine. 

After a couple of weeks I will purchase a fifth spare drive even though they come with a 5 year warranty.
I have learned from bitter experience that in eg. four years time when a drive fails I will send it to WD and perhaps I will get the same model back or perhaps a substitute that is not compatible.  This applies to all HDD manufacturers.  Need to buy the spare drive now while it is available!

Now all I need to do is copy from the degraded DNS-343 to the new DNS-343.
I recall the DNS-343 has an internal backup facility in the installed applications.  My LAN is 1Gbit, I roughly compared the time to do file copies using a desktop versus using the DNS-343 internal copy.  As you would expect the direct process is at least several time faster and does not tie up a desktop for hours.  Another issue with massive file copies is if the desktop has low installed ram then it cannot hold the entire file list in its memory so it only copies the corresponding files and all that without flagging an issue.  Hopefully the DNS-343 has that covered.

It's worth mentioning how to copy from one DNS-343 to another as its not really intuitive.
Log into the Applications of the destination DNS-343 and you will be in the Scheduling tab
Enter your network username & Password if required.
Select "Folder"
Enter the URL of the source drive and click on "Test"
(It is assumed that the source drive is in a different DNS-343 to the one logged on to. 
If not then click on the "Local" button to set the URL)
The URL is a UNC path and for me the syntax is "\\ZEN\Volume_1-1\"
The "Save to" path is local so click on the "Browse" button to set.
For me I select "Volume_1"
That's it.  Click on "Save settings" and an entry will appear in the schedule list below along with a button that allows you to manually start the copy process immediately.

I have elected to go the RAID 5 method again rather than discrete drives due to the inconvenience of a number of logical drives.  To date I have lost a drive three times now in my DNS-343's running Raid 5. I do make copies on USB drives the frequency of which varies by application.  I have never had a problem with data loss where I have had to go back to my backup drives.  If I had discrete volumes then I would lose currency each time I lost a drive.

I have a couple of theories about why people have issues with Raid 5.

1/ As the heavy weights of this forum have said many times, do a full backup first.
A cowboy approach would be to say I'm not spending the money on backup drives, I'll just fix the raid.

2/ The wrong drive is pulled out and replaced/regenerated resulting in all data lost.
Mostly the DNS-343 reports the correct faulty drive but you can be misled:
The S.M.A.R.T. test looks at counters, once that have been incremented for any reason then it will report that drive as bad, especially if you don't look there very often.  Remember S.M.A.R.T. is dumb.
Raid 5 allows you to remove a physical drive without affecting the data.
The smart thing to do is to physically remove the suspect drive, and see if you can still read your data.  If you can do this then you have correctly identified the faulty drive.  If not then play Russian Roulette until you find it.  Of course you don't remove or add drives with power on.

Another issue is where a direct replacement drive is not used and the replacement is not compatible, this just might lose all of your data when you regenerate.

I would be interested to hear any feedback about experiences with the S.M.A.R.T counters.
I leave my drives running 24/7  any thoughts on this?

[melvynadam]

I have a couple of theories about why people have issues with Raid 5.
...
2/ The wrong drive is pulled out and replaced/regenerated resulting in all data lost.
Mostly the DNS-343 reports the correct faulty drive but you can be misled:
The S.M.A.R.T. test looks at counters, once that have been incremented for any reason then it will report that drive as bad, especially if you don't look there very often.  Remember S.M.A.R.T. is dumb.
Raid 5 allows you to remove a physical drive without affecting the data.
The smart thing to do is to physically remove the suspect drive, and see if you can still read your data.  If you can do this then you have correctly identified the faulty drive.  If not then play Russian Roulette until you find it.  Of course you don't remove or add drives with power on.

The SMART test reports the drives' serial numbers. So surely that's the way to correctly identify the drive?

[pbulley]

I am in the process of building 2 NAS boxes, both DNS343s, to replace my collection of USB discs. I upgraded both to v1.0.5 firmware. Populated with 1x 500GB, 1x 1TB and 1x 1.5TB discs, all from external USB drives I disassembled. I wanted to add the 4th disc in each so purchased 2x ST1500DM003 drives. These were recognised by each DNS343, but both failed to format after getting to < 10% and apparently stopping. I swapped these over and got the same result. I then fitted them to a couple of my now spare USB enclosures, then mounted and NTFS formatted in Windows 7. All worked fine. I therefore concluded: not compatible! Went out and bought 2x WD Caviar Green drives, added to DNS343s, formatted, setup shares, etc, all fine. Whilst annoying, I'm not heartbroken and will use the Seagate drives to replace smaller internal discs in my 'server' PC. I guess there's something common to these Seagate drives? Seems too much of a coincidence really.

[JAylmer]

I suspect the S.M.A.R.T. error counters in the 2x ST1500DM003 are now max'd out or very high error values when you attempted the DNS-343 format/array build.

The counters are sadly falsely incremented and cannot be reset. 
Do you have any way of having a glance at the SMART error counters to confirm?
If not convenient then no probs.
Could you please give us the model of the WD green drives that worked OK for future reference.

The SMART test reports the drives' serial numbers. So surely that's the way to correctly identify the drive?

My main concerns about SMART have just been listed  in your thread: "RAID 5 (Degraded)" but drive now shows status as "Normal"

[ziden]

I am having similar problems with ST1000DM003.
It's the only disk i am inserting now.
After formatting reaches 99% it stops with error 17 without any description and googling did not gave much results other than questions form other people facing the same. 

I have started form firmware 1.02 where i first faced this, than it was 1.06beta1 with formatting error without even a number now. I am a bit desperate. Has anyone managed to achieve results with this Seagate ST1000DM003 at all ?