Print

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

[melvynadam]

I'm running a DNS-343 with f/w 1.05 and four identical Seagate Barracuda (7200 1.5TB ST31500341AS) drives configured as RAID5.

My NAS has been suffering from issues for a while and some have been worked through on this forum but none were completely resolved.

For a long time I thought the issues were caused by a specific drive. So I bought an identical unit and replaced the drive. For a very brief time, that seemed to resolve it but then the unit went back to complaining that "VOLUME_1 is degraded" every day. Things mostly worked although when streaming media, I occasionally needed to restart the box. Clearly far from optimal. Tonight, when I restarted the box, the entire array has simply vanished  

When I login through the web control panel, I see the following:

Status>Hard Drive Info:

Slot	Manu.	Model	Serial Number	Temp.	Status
1	Seagate	ST31500341AS	9VS25J7Q	0°C / 32°F	Abnormal
2	Seagate	ST31500341AS	9VS24HVH	43°C / 109°F	Normal
3	Seagate	ST31500341AS	9VS3XRCM	44°C / 111°F	Normal
4	Seagate	ST31500341AS	9VS3SVYJ	40°C / 104°F	Normal

Maintenance>Disk Diagnostic:

Slot	Manu.	Model	Result
1	Seagate	ST31500341AS	Pass[ 2012/09/20 04:19:35 ]
2	Seagate	ST31500341AS	Pass [ 2012/09/20 04:16:15 ]
3	Seagate	ST31500341AS	Pass [ 2012/09/20 04:16:22 ]
4	Seagate	ST31500341AS	Fail [ 2012/09/20 04:15:50 ]

So, how can I be sure which drive has the problem? And, more importantly, how can I get my RAID5 array back? FWIW, I've read the two similar threads (here and here) but neither contains answers relevant to my present situation.

[JavaLawyer]

Have you had any luck or learned anything new to identify the "true" faulty HDD?  Or do you think Murphy visited your NAS and there really are two failing HDDs?  

If I could offer more insight here I would.

[melvynadam]

No luck. My box has been switched off for days. All the time I'm hoping that the "replies" count to this post will miraculously keep up with the views 

My main question is this: if I click "Re-configure all existing hard drive(s)" and select RAID5, will it format all the drives and kill all existing data on them, or will it recreate the array I had (assuming three drives in there are ok)?
If the latter, I'll be pulling out the drives instead to attempt to pull the data off using a 3rd-party piece of software ("Runtime" as discussed here: http://forums.dlink.com/index.php?topic=6144.0).

[melvynadam]

Does anyone know what "Re-configure all existing hard drive(s)" does?

[JavaLawyer]

Does anyone know what "Re-configure all existing hard drive(s)" does?

DNS-343 manual:

• Re-configure All Existing Hard Drive(s):  Click Re-configure All Existing Hard Drive(s) to setup the RAID configuration. Note: All data will be lost when changing the drive configuration.

I believe you may be confusing re-build and re-configure.  The re-build functionality is used to repair/restore a failing array, while the re-configure functionality is used to create a new array (or overwrite an existing array).

[melvynadam]

DNS-343 manual:

• Re-configure All Existing Hard Drive(s):  Click Re-configure All Existing Hard Drive(s) to setup the RAID configuration. Note: All data will be lost when changing the drive configuration.

I believe you may be confusing re-build and re-configure.  The re-build functionality is used to repair/restore a failing array, while the re-configure functionality is used to create a new array (or overwrite an existing array).

And I believe that "re-configure" is a poorly chosen name for this function, and that the documentation is sufficiently vague as to leave open the possibility that it does what I'm looking for.
I wasn't confusing it with rebuild. If the only function performed by this button and its underlying code is to configure an array, why call it "re" anything? In the event that there is no current configuration, what is it "re-configuring"? The documentation says that data will be lost "when changing the drive configuration". I'm not attempting to change the configuration, just restore what was already in existence. So, perhaps the data won't be lost?

[JavaLawyer]

I agree that the choice of nomenclature, i.e. re-configure vs. re-build, is quite subjective and open to multiple disjointed interpretation.

Nonetheless, I am fairly certain (someone please correct me if I'm wrong) that the re-configure function is a two step procedure that formats the HDDs and then creates the array on the fresh volumes, while the re-build function attempts to restore existing data.

Please visit the Support link (online help) in the DNS-343 web UI and navigate to Advanced > RAID. The support language states:

"Raid
Choose from four different hard drive configurations in this section. Begin by pressing the 'Re-configure All Existing Hard Drive(s)' button. Note that all data will be lost during this process."

As ambiguous as the name of the function might be, the description and warnings are clear.

[JavaLawyer]

As a follow-up question, do you have a backup of your data?

[melvynadam]

As a follow-up question, do you have a backup of your data?

Kinda. Which is why I'm keen to get the drives back up and running rather than restore backups.

My DNS-343 is for two purposes:

• shared location for the whole family to put documents, photos, backup files, and music
• media server for movies and TV

The important stuff (bullet #1) comprises 700GB and is all backed up to Crashplan so it's in the cloud and can be restored whenever I need.

The less critical stuff (bullet #2) was around 3TB and is not backed up at all. It represents quite a lot of material sourced from multiple friends and efforts over a great deal of time. Backing it up was deemed too costly and time-consuming so I haven't done so. Losing my kids' old episodes of Magic Schoolbus isn't as painful as losing family photos from the last twenty years would have been.

So, although the most vital material is backed up I would still really like to get anything available from those drives that I can.

[JavaLawyer]

Although not a guaranteed solution, many users were able to fully recover their files from non-cooperative RAID arrays on various ShareCenter models (including the DNS-343) using the third party software products mentioned in the following thread: DNS-343 - Data Recovery (Windows PCs).

[melvynadam]

using the third party software products mentioned in the following thread: DNS-343 - Data Recovery (Windows PCs).

Thanks, I mentioned the same tools above:

using a 3rd-party piece of software ("Runtime" as discussed here: http://forums.dlink.com/index.php?topic=6144.0).

Yesterday I downloaded Seagate's SeaTools and ran the "Long Generic" test on the original drive to fail in my 343.
That drive failed the test and is still under warranty so I'll be replacing it. Next, I'll test the others and see if I was just supremely unlucky having two drives fail simultaneously in a RAID5 array. After I've run the tests, I'll be using Runtime to try and see if any data can be pulled off, then I'll get Seagate to replace any faulty drives, and finally I'll be able to get my NAS back up and running again.

Total downtime is unlikely to be less than two months. Think I might give RAID a miss this time 

[melvynadam]

I've got a nagging suspicion that the NAS is at fault. Other than by putting yet more shiny new drives in the firing line, I'm not sure how to test that theory. Any ideas?

[JavaLawyer]

I've got a nagging suspicion that the NAS is at fault. Other than by putting yet more shiny new drives in the firing line, I'm not sure how to test that theory. Any ideas?

If the Seagate HDDs did in fact fail, I don't see how the DNS-343 is capable of killing the a HDD, which I suppose (as you stated earlier) would make you "supremely unlucky".

[melvynadam]

If the Seagate HDDs did in fact fail, I don't see how the DNS-343 is capable of killing the a HDD

I'm nowhere near as knowledgable as you on the D-Link products but I've reviewed all of my actions regarding these issues and concluded that Slot #1 in my DNS-343 is the common denominator. All three drives that have failed have spent some time in that slot. So, either it's an amazing mistake that three HDDs have failed in such a short time of each other, or there's a short or some other hardware issue in that slot.

[JavaLawyer]

. . . I've reviewed all of my actions regarding these issues and concluded that Slot #1 in my DNS-343 is the common denominator. All three drives that have failed have spent some time in that slot. So, either it's an amazing mistake that three HDDs have failed in such a short time of each other, or there's a short or some other hardware issue in that slot.

A hardware failure is something I haven't considered until now, but your logic makes perfect sense. I think I already know the answer to this (based on when you registered in this forum), but have you had your DNS-343 for more than one year? If there is in fact a hardware failure, you're best recourse is to try to get a replacement unit from D-Link.

A phone call to their technical support may not help, but it can't hurt.