Print

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

[melvynadam]

I have four Seagate 1.5TB ST31500341AS drives configured with RAID5. The box has been running flawlessly since July 2011. I currently have f/w rev1.03 and "Used Space:  3781079 MB", "Unused Space:  642025 MB".

Suddenly on Wednesday night while streaming a show from the NAS via a media streamer, performance became erratic and I checked the admin pages to see the fourth drive listed as "Abnormal" and the "Sync Time Remaining" defined as "Degraded". I tried to shutdown but the system wouldn't shutdown properly so I pulled the power.

I came back to it tonight and plugged it back in. It still shows "Sync Time Remaining: Degraded" but now the drive shows as "Normal".

I've ordered a new ST31500341AS in case the problem really is the drive but I have some questions that I thought the experts here migh be able to help answer:

• If the drive now shows as "Normal", should I click "Re-configure All Existing Hard Drive(s)" and see how long things last? Or is it risky in case the drive is faulty and the rebuild fails midway?
• I've ordered the new Seagate drive now but I'm curious about whether I needed to buy the exact same model. Would I have been fine with any 1.5TB drive?
• When the new drive arrives, what are the steps to get my system back up and running again?

Thanks so much for taking the time to read this far.

[melvynadam]

50+ views but no replies. I know this is an enthusiasts' forum and nobody owes me anything but can anyone provide any insight at all on my situation?

My new drive arrives tonight. When I receive it, do I simply swap out the old for the new and click "Re-configure All Existing Hard Drive(s)"?

[melvynadam]

Some more info:

• the 4th drive fails a SMART test in the maintenance tab
• after a cold restart my data was all available for about 12hrs before things went flakey again

This thread implies I should remove the drives, update my firmware, replace the drives, and pray. Is that a good plan?

[JavaLawyer]

If you haven't already done so, you should create (and maintain) a full backup of your data.  RAID-5 (or any form of RAID for that matter) is not considered a backup, but only provides redundancy: DNS-343 - Data Backup vs. Redundancy

[melvynadam]

Thanks but I can't access the data. Besides, I do have a full backup of the mission critical material.

[melvynadam]

As far as I can tell, there are three tasks I should do:

1. Replace the supposedly faulty hard drive
2. Rebuild the RAID5 array
3. Update the firmware to 1.06

I think the order should be as listed above but maybe it should be 3,1,2. Anyone?

[JavaLawyer]

I'm glad that you have a backup of your critical data. Moving forward, if you are looking at the DNS-343 as a single-source for your original and backup data, you should consider using a non-RAID solution. This advice applies to any brand NAS.

Page 25 of the following updated DNS-343 manual (for FW 1.04) describes the process for auto-rebuilding RAID for a failed HDD: DNS-343 Manual WW.

The FW update history does not indicate any changes between v1.03 and v1.04 that impacted/improved RAID handling, nonetheless, I can't definitively state whether updating the FW will help, hinder, or have no impact on this process.

[JavaLawyer]

For your personal edification, here are the cumulative FW release notes: DNS-343 - Cumulative Firmware Release Notes

[melvynadam]

Moving forward, if you are looking at the DNS-343 as a single-source for your original and backup data, you should consider using a non-RAID solution. This advice applies to any brand NAS.

Thanks. Not sure why that's advisable. Surely, the redundancy inherrent in RAID5 provides some peace of mind so that, in the event of a single drive dying (as mine apparently might have done) you are able to rebuild and carry on as if nothing has happened? BTW, the GUI actually makes a counter claim to the one you're asserting. It says "RAID 1 (Mirroring - Keeps Data Safe)".

Also, I had just reviewed the cumulative release notes and was reminded of a question I asked about 1.04 which was never answered. Can you shed any light?

[JavaLawyer]

Thanks. Not sure why that's advisable. Surely, the redundancy inherrent in RAID5 provides some peace of mind so that, in the event of a single drive dying (as mine apparently might have done) you are able to rebuild and carry on as if nothing has happened? BTW, the GUI actually makes a counter claim to the one you're asserting. It says "RAID 1 (Mirroring - Keeps Data Safe)".

Not entirely true:

• If you accidentally delete a directory on Volume_1 of a RAID-1 array the data is instantly and permanently deleted on both volumes.
• If a virus infects a RAID-1 array, then the virus will instantly propagate to the second volume.
• If a virus infects a RAID-5 array, then there is no clean backup.
• If a RAID-5 array gets corrupted, then all data is lost

[JavaLawyer]

I had just reviewed the cumulative release notes and was reminded of a question I asked about 1.04 which was never answered. Can you shed any light?

Please provide more information on the ramifications of this feature:

• UPnP AV Server updated and DLNA 1.5 Certified

My only real gripe with the DNS-343 is that I often have to perform manual refreshes of the UPnP server to "see" new items from my media streamers. Will this firmware resolve this issue for me?

I haven't personally used the updated UPnP AV Server, and am therefore unable to comment on the improvements other than what's already stated in the release notes. DLNA certification is simply a nod from the Digital Living Network Alliance acknowledging D-Link devices as DLNA compliant.

[melvynadam]

Not true.

Well firstly you're not arguing with me - it's a claim made in the GUI (and the screenshot appears in the manual you pointed me towards).
I agree that RAID is no substitute for backup. A large chunk of my NAS is backed up to the cloud with an automated process and I don't use RAID with backup in mind.
All of your points are correct and valid. Nonetheless, if you have four 1.5TB drives and 3TB of data to store on them, I'd argue you're better off with a 4.5TB RAID5 array than single volumes. As long as you're backing up too, in the event that a drive dies, in theory you're going to make a quicker recovery by replacing the drive and rebuilding than by replacing the drive and restoring from your remote location. I'll let you know if the theory holds up when I get my new drive tonight

[JavaLawyer]

Well firstly you're not arguing with me - it's a claim made in the GUI (and the screenshot appears in the manual you pointed me towards).
I agree that RAID is no substitute for backup. A large chunk of my NAS is backed up to the cloud with an automated process and I don't use RAID with backup in mind.
All of your points are correct and valid. Nonetheless, if you have four 1.5TB drives and 3TB of data to store on them, I'd argue you're better off with a 4.5TB RAID5 array than single volumes. As long as you're backing up too, in the event that a drive dies, in theory you're going to make a quicker recovery by replacing the drive and rebuilding than by replacing the drive and restoring from your remote location. I'll let you know if the theory holds up when I get my new drive tonight

As long as there is an external backup, I wholly agree with you. RAID is a means for ensuring data availability and minimizing downtime. Many users confuse RAID and backups, so I use every available opportunity to get on my soapbox. 

Good luck with your recovery.

[melvynadam]

Good luck with your recovery.

Thanks. Do you have a recommendation between these two options?:

• Replace drive, Rebuild array, Upgrade firmware
• Upgrade firmware, Replace drive, Rebuild array

[JavaLawyer]

Thanks. Do you have a recommendation between these two options?:

• Replace drive, Rebuild array, Upgrade firmware
• Upgrade firmware, Replace drive, Rebuild array

As I said earlier, "The FW update history does not indicate any changes between v1.03 and v1.04 that impacted/improved RAID handling, nonetheless, I can't definitively state whether updating the FW will help, hinder, or have no impact on this process." Unfortunately, given my uncertainty I'm not in a position to recommend which course of action is best suited to your situation.