Print

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

[Absynthe]

I was reading the FAQ and manual and it said that if my RAID1 is degraded, that means a drive is going bad. Ok.

It also said that if I look on the case, it'll tell me which drive because that drive will have an amber light on instead of blue.  Hmmm.

Well, according to my disk tools, the RAID1 is degraded, but both drive lights are blue.  So... which drive is bad?  Under AutoRebuild Configuration, it's listing only one of my 750gig drives.  Is that the good one?

Current RAID Type :
   Volume_1 : RAID 1 (Degraded)

[gunrunnerjohn]

Obviously, that's the good drive.  My next move would be to take both drives out and test them externally.

[Absynthe]

Thanks for answering, I was thinking that was the good drive too.

How would I go about testing them?  My understanding is that they can only be tested on a Unix OS?  I have all WinXP laptops. 

I just ordered an external enclosure, figuring worse comes to worse, I'll reformat the "bad" drive as NTFS and test it that way.  See how many bad sectors come up.

[gunrunnerjohn]

If you download and run the disk manufacturer's diagnostics, you can do a full non-destructive surface scan and SMART test.  This will leave any data intact on the good drive.

[Clayton]

Can you do this while the drives are still installed in the DNS-323 unit?

[gunrunnerjohn]

No, you need to do this connected to a desktop that can boot the diagnostics disk.

[fordem]

I was reading the FAQ and manual and it said that if my RAID1 is degraded, that means a drive is going bad.

I just noticed this - and thought I'd comment on it.

A degraded RAID array is one that is no longer redundant, and in the event of a disk failure, will/may cause data to become unavailable.

It does NOT mean that a drive has failed, is failing or is about to fail.

In the case of a DNS-323 with RAID1 a degraded array could be something as simple as the data on the drives being out of sync for whatever reason (a power failure will do it), and both drives could be fully functional.

This is by no means unique to the DNS-323 and I have seen it happen with other servers.

[gunrunnerjohn]

Well, it could also be a bad drive.  I always worry about trying to resync them without knowing, just in case the box gets confused and really screws things up!

[eyeoh]

Obviously, that's the good drive.  My next move would be to take both drives out and test them externally.

Hello.  New here.

My experience does not coincide with the above quote.

I had encountered this same situation recently - DEGRADED RAID1 on my DNS-323.  FWIW, I have email alerts enabled, and within the email which the 323 sent me, the drive that was noted as FAILED (in my case, LEFT) is the same drive that is listed on the "AUTO-REBUILD CONFIGURATION" of the 323's Tools-> RAID web page.

DNS-323 web page TOOLS->RAID:

RAID CONFIGURATION :
 Current RAID Type :
      Volume_1 : RAID 1 (Degraded)

AUTO-REBUILD CONFIGURATION :
 Vendor      Model               Serial Number     Size
 Seagate    ST3750640AS    5QD1Nxxxx         750 GB


Email message:

Left Hard Drive Has Failed

Sincerely,
Your DNS-323


DNS-323 web page STATUS->DEVICE INFO:

PHYSICAL DISK INFO :
Slot    Vendor      Model               Serial Number    Size
Right   Seagate    ST3750640AS    5QD1Mxxxx          750 G    
Left    Seagate    ST3750640AS     5QD1Nxxxx          750 G



So, from the above, I gather that the drive listed in the AUTO-REBUILD CONFIGURATION during a Degraded RAID1 incident is actually the FAULTY drive - not the good one.

With that said, though, the lights on the front of the DNS-323 were both blue - neither amber.  I enabled a rebuild, and it never completed, again indicating the DEGRADED RAID1 scenario above.

I attempted a disk scan using D-Link's own utility found in TOOLS -> DISK TOOLS but it would fail prior to completion.

I powered down the DND-323 (it was very very warm at this point) and left it off overnight.  Powered on in the AM, enabled a RAID rebuild, and it completed successfully.  I'll monitor it over the next few days.  Should it fail once more, I'll take the drives out and will perform individual drive scans using SeaTools and RMA if necessary.

HTH.

[EinHexenMeister]

well where to start

i have a similar situation where 2 of my 4 323s are degraded

i had a power failure and one (323#1)showed the degraded error ... this unit had a problem about a year ago when i started to use it 1st ... i also remember in the old days that one of the leds went amber

so i do more tests after the power failure ... try to re-seat the drives with the power off and managed to plug one drive in backwards (it was 2am), with the connectors to the outside and power it up and run disk check, so no hw harm. now this unit (323#2) is also degraded.

now the questions are

1) if there is no amber light ... would this mean that both drives are good and working

2) according to earlier conflicting posts ... the drive indicated by it's serial number in the tools -> raid page in the auto rebuild configuration is now the - a: good one - or - b: bad one - how do i know for sure

3) unfortunately i do not have a 100% backup of the data on both units ... if i copy all data from these 2 degraded raid units to external drives how confident should i be that all the copied data is good ... there is no way that i can check 1TB+ of data by hand for integrity

4) do the lights still change colors with rev 1.08, and if  ... when ... can this be verified, by using a forced failure

5) is there any documented way which indicates the good drive and the bad drive after a degrade event


i have to say the d-link docs are lacking majorly in details pointing out ... what's left and right, front view or ejection lever view, back ... i figured out it's front by now

also the serial number in the auto-rebuild config section is completely useless without designating it as the failed one or the good one

somehow i feel like being in las vegas at a dns-323 slot machine pulling the manual rebuild now button

there is no way that i will ever be using raid on a 323 with the lack of sufficient docs

yes ... we all know what's a raid but who knows what the 323 is doing once it fails ... it can do anything cause the docs allows it to

thanks and cheers EinHexenMeister

[159159]

There's my story too.
DNS-323 firmware 1.09, Cyberpower UPS

   Checking web interface I found that RAID-1 is degraded, both drive lights was blue, “physical disk info” under “Status” > “Device Info” shows both drives are online. Logs didn't show an event related to RAID status or HDD failure.
   After rebooting DNS-323 I rebuilt RAID-1 and used NAS with no problems for about a week. Then one day found RAID-1 is degraded again. Both drive lights was blue like the first time, but “Status” > “Device Info” > “physical disk info” showed just one Left HDD and log says that “HDD1 Hard Drive Has Failed”. Means HDD2 is Left drive and HDD1 is Right one -confusing. After restart, DNS-323 showed both HDD in “physical disk info” again, but one with 0GB on it. Still no amber light on front panel.
   Problem is I don't know when HDD is failing or failed if there's no indication on the unit and that's very frustrating. Log says HDD1 has failed and it's not clear is it left or right. Yes I can check under Device Info, but will it would be more clear to use same device names in the log too?
   HDD replacement arriving tomorrow and I'm preparing myself for the worst...

[gunrunnerjohn]

You might consider sticking a label on the front of the drives with their serial numbers.  That way you can correlate the status with the physical drive without guessing which one it is.

[m2k3423]

After power recycling, the drive seems "recovered" from the problem because the sector remapping happened. I will suggest you have the drive replaced soon.