Print

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

[daz10000]

Apologies in advance for this rant, but this <many expletives withheld> little device has wasted so much of my time. I've maintained my own linux RAID stack for years and finally decided maybe six months ago to move a lot of my key storage over to the 343.  I have two 323s and they have worked flawlessly for years. The 343 seemed like a nice step up and I figured it would be low-maintenance. I couldn't have been more wrong. Without going through a blow-by-blow description, this system has just been an endless series of headaches. I have never once in over a dozen attempts to reformat it, managed to get a successful RAID5 volume. It would pretend to format things then reboot and not be able to find what it had just created. It would only recognize three drives sometimes. There would be a perpetual flashing light on one drive, but it would be a different drive from time to time. I sent it back after a lot of unhelpful troubleshooting and they sent a new one which had the same issues. I gave up for a while and just lived with two separate volumes 1 with three drives and one with the fourth drive, but I didn't want anything significant because I didn't trust it. It would go off-line from time to time for I would visit the console and find just the single drive available. A reboot would temporarily restore the three drive volume. Finally in frustration after a good run fixing a bunch of other things I decided to give it one last try. I carefully reformatted all of the drives separately on a desktop machine so there were no existing partitions to confuse it. I flashed the device one more time so that everything was clean. At this point I noticed the front OLED panel has apparently failed despite the device not doing anything useful for six months. Nothing has been able to restore it. I was delighted to see my four 500Gb Seagate drives and attempted one last RAID5/ ext3 format. It ran really smoothly right up until the last few seconds and then mysteriously failed with error 127.  I can't find any documentation anywhere on this mysterious error message. Worse still, once I rebooted, the device just insisted that I had the drives inserted in the wrong order, and as I'm sure you know is nothing you can do from the user interface at that point. I removed the bottom drive and managed to get it to at least come up and talk to me. Reinserting the fourth drive that tells me that there are four drives available, but the wizard will only let me build a raid volume with three of them. I have probably wasted a total of 3 to 4 days on this machine, plus shipping to replace the hardware and now I have a very expensive black brushed metal paperweight. I'm just total loss to understand what went wrong.  Dlink made some great devices on this seems like a small upgrade, but I'm not sure I will ever touch another dlink product again after this experience. Ironically these four discs came out of a working linux RAID that really never gave me any trouble until the power supply on the motherboard failed and I thought it might be more convenient to have storage and server physically separated. Big mistake.

Thanks for listening and if anybody knows what error 127 is or where these kinds things must be documented, or good uses for a completely <$%^&*>ed NAS device, let me know. I'm guessing it's fully #$%^ed and it has probably cooked itself and four of my drives at some point which is a pity,

Darren

[JavaLawyer]

What firmware version are you using? Are your HDDs advanced format?

[daz10000]

No advanced format drives.  These have all worked in a previous internal linux array so I don't expect problems from the drive.  The most recent experience I talked about was with 1.05 but most of the early testing was 1.04.   I left the unit on over night to see if would recover anything after my last post and the front OLED never came on,  the missing 3 volume raid never returned but the unit made a pretty horrible clunking sound periodically through the night (sounded like maybe drives resetting).   There was nothing in the logs except periodic "fan failed" messages.  I suspect the unit at some point failed to cool itself and it has done hardware damage.  The OLED only recently stopped turning on at all, but when it was working it didn't really provide much information about what's going.  One of the problems with the design of the unit is very little information about what happens.  I'm sure it would be informative to see detailed logs but nothing appeared on the web logs

[JavaLawyer]

The most recent experience I talked about was with 1.05 but most of the early testing was 1.04. . . .There was nothing in the logs except periodic "fan failed" messages.  I suspect the unit at some point failed to cool itself and it has done hardware damage. . .

Firmware v1.04 generates an erroneous "fan failed" message that was resolved in firmware v1.05.

[JavaLawyer]

The failure of the OLED to turn on is indicative of a hardware failure (or perhaps a severe firmware corruption?). Did you try flashing the firmware to the factory defaults by holding the reset button on the rear of the DNS-343 for > 8 seconds? This procedure should not impact your existing data or HDD configuration. If you choose this route, I recommend saving a configuration file (through the web interface) to quickly restore your user settings.

[daz10000]

To be clear all the problems I have had started long before oled failure, but that's maybe a symptom of the machine either damaging itself or corrupting itself.   I started off the last clean test by flashing the device, and erasing all partition info on the drives so it could have a fair test.  That test failed (at 100%) with the error 127.  I applied firmware 1.04 for that test.  I then tried the 1.05 beta in hope it might resolve things, but it didn't do any better.  The oled has  been off for the last week or so.   I suspect the hardware has damaged itself but it continues to respond ok to the web and is otherwise ok (except that it doesn't really behave like RAID).  At this point I have an expensive paperweight, no recourse to getting it fixed and it might have done the drives harm if they overheated at some point.  I've RMAd one of these at some expense and am reluctant to do it again unless there is some confidence that the product can work properly.  I'm not getting a sense that this whole design is very reliable.

[JavaLawyer]

I have two DNS-343s operating without issue for ~3 years now. I firmly believe one of the reasons I haven't encountered any substantial issues is due to the fact that all 8 HDDs are configured as Standard Volumes. From reading these boards over the years, the RAID implementation seems problematic for a number of users and for a myriad of indeterminate reasons, some of which I'm sure are related to the HDD make/model and on-board software-based RAID controller. That said, and although four independent volumes may not be ideal for most users, I would hazard to guess that simplifying your HDD configuration should resolve your data access and stability issues moving forward (i.e. assuming your DNS-343 and HDDs are undamaged).