"Disk Failed", then "Disk Unplugged" errors

[sgalbincea]

Same issue here: TS-1263U-RP on firmware 4.3.4.0597 - lost four of 12 drives over the weekend after seeing the exact same behavior for a few weeks now. Support ticket opened after waiting on hold for an hour then finding that the voicemail box for support was full. Not happy, and definitely not buying another QNAP product for a client again. Thankfully this is a secondary copy of our Veeam backups here, so I am not out any data if the whole thing rolls over.

[psycho_asylum]

TS-563 here, disk 5 dropped out and started rebuild then kept coming back and knocking disk 4 out. Of course I'm out of warranty so this will probably be my first and last QNAP device. I paid almost $600 for this thing and haven't changed drives but once, so it's not wear and tear. My data is safe but I'm quite annoyed that I need to find something to replace it. There's nothing really comparable to this with the same horsepower at this price point.

[thesero]

Hi All

I have a QNAP TS-431P2 bought less than 2 months ago, with the latest firmware.

Few days ago I have put 3 disks inside, one 2.5'' seagate 500GB and 2 WD Red 3TB. All 3 were successfully used before for more than 2 years.
Now, after 3 days, got some warnings, then suddenly I have only one WD drive active, the other two are no more shown in the qnap disks dashboard.

Not having used the NAS enough time to lose data, but 2 disks is quite some money, and then, should I trust this device for keeping my important data?
What is wrong?

Thanks,
Sero.

[Pateriko]

Same issue here, TS-831X, Disks 1, 4 & 7 all failed, degraded, unplugged, within 2 minutes of each other.

Different size & Brand disks, all in seperate RAID1 arrays with identical drives.

Disk 1 & 2 are Seagate Ironwolf 8TB's, Disks 3 & 4 are WD Red 8TB's & Disks 7 & 8 are Seagate Ironwolf 4TB's.

Slots 5 & 6 are empty so this has degraded all 3 of my volumes.

[shaneking]

I have been experiencing this sort of issue for some time now with two different QNAP Nas devices. The first TVS-463 had 4 HGST 4TB hdd installed and the reseller replaced the drive under warranty. It didn't resolve the problem. He then arranged a warranty replacement of the NAS with QNAP. About 8 months the same issue happened Drive 3 reported as failed but when the SMART diags were run on HDD3, it showed the drive as healthy. I lodged a support request with QNAP, they again replaced the device with a TVS-473.

It looks like an issue with QNAP hardware.

I know this doesn't help your situation but being aware others are seeing this, may at least encourage you to lodge a support request with QNAP

[garyschmidt78]

Oh no, sorry to see this thread. I have the same issues in my server that reared its ugly head last week. My TVS-473 had Disk 4 sputter (green light went off and then on again). System rebooted and OS went to rebuild Disk 4 and when it was about 70%, Disk 4 light turned off again. Then I opened a ticket on Friday last week and then on Monday I was instructed to run Western Digital Lifeguard on Disk 4. Drive PASSED. Sent screenshot to support ticket and I was than asked for my original invoice on Tuesday for warranty verification. Sent Tuesday night. Haven't heard a peep back in three days from support ticket despite a followup email on my ticket number. After reading this thread, it indicates there may be a serious problem with QNAP servers since I'm not alone. I was suspecting a power supply voltage issue, but perhaps there is a SATA backplane ASIC that is faulting out and removing/unplugging otherwise healthy Disk Drives.

[P3R]

After reading this thread, it indicates there may be a serious problem with QNAP servers since I'm not alone.

You have a short history of being an alarmist around this forum so let's analyze the claim that there "may be a serious problem with QNAP servers".

This thread at the time of starting to write this post had a total of 160 posts. Many users have posted several times on the same issue. Some are false positives that really have disk issues (the by far most error-prone component in a NAS) so a very rough estimation (done by me) is that probably 50-70 users in total are affected. Let's round upwards and say 70. At least one forum user represented a very large number of units where something like 30-50 units are affected (that guy I really have a lot of sympathy for ) so let's add 50 and arrive at 120 single units. So is 120 units or even the double number with this problem a serious problem given how many units Qnap sell in total and how many units that are working for other users? I don't think so.

To a user with a single affected unit it's a 100 % fault rate and that single failing NAS may be a disaster for them (especially if they haven't followed the standard recommendation from all knowledgeable users to regularly backup any data considered valuable on a another system). But to keep things in perspective, that doesn't mean that Qnap have a serious problem or that it will affect even a fraction of their other customers. As the cases reported here (or even the double number) is a very low precentage of their total sales volume. The risk for being affected as a customer is still very low.

Having followed this thread I think it's clear that the TS-X63 model have been affected by some kind of production issue. TS/TVS-X73 may also be slightly affected but in relatively very low numbers.

Complicated technical gear sometimes fail, that's the reason there are warranties. I agree that it's inconvenient and very annoying when it happens but it's not really the end of the world as you make it sound.

[theyeti]

Unfortunately I have to join in here - my TS-831X started "unplugging" drives earlier this year. Initially I swapped drives for new ones, but the issue kept returning. As of now, 3 out of 8 bays "unplug" drives a minute after fitting. The HDDs in question are brand new HGST Deskstar NAS and Toshiba N300 which all test ok when attached to a PC.

I opened a ticket two months ago, but all I received was automatic replies, the last one suggesting to consult some Wiki and update firmware. Then the ticket was closed. I managed to find a phone no to call - in the Netherlands, as they do not have one in Austria and only a 900 toll number in Germany. At least, the lady I spoke to was friendly and appeared sensible, mentioning a "backplane issue". I mailed a log dump, so let's see what happens next.

I seriously hope there is a reliable fix for this issue... (And always remember to back up your data!)

[bookpop]

You can also add me to this list, having experienced the issue about 18 months after initial purchase of my TVS-673. My setup is a RAID 5 with four drives (1,2,3,4) and a hot spare (5). Initially, the problem presented itself with Drives 3 and 4 ejecting. After this happened three times in an hour, the RAID crashed, is "not active", and recovery does not work. My hot spare (5) now shows as regular data drive, and my regular data drives (3,4), show as Spare and Free, respectively.

Running md_checker, it appears my RAID has somehow gotten split:

Code: Select all

Welcome to MD superblock checker (v1.4) - have a nice day~

Scanning system...

HAL firmware detected!
Scanning Enclosure 0...

RAID metadata found!
UUID: d574cff5:68eee6bd:2d3af176:989606db
Level: raid5
Devices: 4
Name: md0
Chunk Size: 512K
md Version: 1.0
Creation Time: Aug 5 08:10:46 2018
Status: OFFLINE
===============================================================================
Disk | Device | # | Status | Last Update Time | Events | Array State
===============================================================================
5 /dev/sdb3 0 Active Aug 5 08:10:46 2018 0 AAA.
7 /dev/sdf3 1 Active Aug 5 08:10:46 2018 0 AAA.
3 /dev/sdc3 2 Active Aug 5 08:10:46 2018 0 AAA.
-------------- 3 Missing -------------------------------------------
===============================================================================


RAID metadata found!
UUID: 13c2826f:5af93bc5:4f0e5af6:f109b81c
Level: raid5
Devices: 4
Name: md2
Chunk Size: 512K
md Version: 1.0
Creation Time: Feb 23 19:27:02 2017
Status: OFFLINE
===============================================================================
Disk | Device | # | Status | Last Update Time | Events | Array State
===============================================================================
-------------- 0 Missing -------------------------------------------
4 /dev/sdd3 1 Active Aug 4 21:21:52 2018 397576 AA.A
-------------- 2 Missing -------------------------------------------
6 /dev/sda3 3 Active Aug 4 21:02:59 2018 397561 AAAA
===============================================================================

All drives are Seagate ST3000VN000-1H4167. I successfully performed a Fix All Long test on the drives using the OEM software. They also passed all SMART tests. I am very much convinced the drives are not faulty. In the meantime, even after adding another drive in Bay 6, I cannot recover the RAID.

In playing around with this some more, I am able to consistently reproduce the spontaneous ejecting issue. If I run a bad block scan on Drive 3, at about 9%, both Drives 3 and 4 eject. To re-recognize the drives, I have to physically eject them and re-insert them. Even then, this only works about 25% of the time. The rest of the time I get the "clicking" on both drives, and they do not show as inserted. I agree with other posts that it almost seems like a power-related issue from within the NAS.

Anyway, I contacted QNAP support almost 3 days ago and have not yet heard a response. At this stage, my primary concern is my data. Although I have about 75% of it backed up on Glacier, the remaining 25% was not. I also don't want to perform a restore of any of that data until I am convinced that my current RAID is trashed and that the NAS hardware is stable. So, I am left unable to access my data until I get some resolution to this. The secondary concern, of course, is whether I am going to need to RMA my unit for this issue and how long that may take.

If anyone has thoughts on repairing the RAID (even temporarily to allow for a remote backup), I would sure like the advice! That would at least bide me some time, as I could temporarily restore select files to a new location.

[compusger]

Well, even when i hate to, but i have to join the club.

TS-1ß79 Pro, Disks 9 and 10 are getting unplugged after a while...
Checked HDDs in another slot, no issues, smart test and so on are fine.

My Qnap is now 3 years old and guess what ?
No Spare parts available any longer.

I find it interesting though, that i have to make a backup of my NAS that has Raid 6 enabled to be as safe as possible...
Anybody any idea, where to get the Backplane replacements ?

Greetings from China

[saliwok]

+1 here.

TS-563 just lost drives 4 & 5 again. First time was about 4 months ago, and I was able to recover then. Not having much luck this time.

2 questions from a NAS noob -

1. Has the "unavailable" part ever been identified? Surely it exists somewhere in the world.
2. Is it possible to put the drives into a new unit to recover any data?

[dolbyman]

1. if it is a custom made part ... probably not that easy
2. Check the migration guide to see units that would work, also avoid this by having backups at all times

[saliwok]

1. It almost feels like a mechanical issue - like a fitting getting loose over time and losing a connection. Whether custom or not, it can either be reproduced or poached off other units. I guess my biggest concern is that the good people at QNAP just can't be bothered to help out.

2. Thanks for the tip on the migration guide. And yes, backups would solve most of the problems on this forum.

[compusger]

Hey there,

Answer to 1:
well i had tickets open and many emails with the Qnap support in Shanghai.
In my case, i pushed as i don't want to accept no for an answer.
The final agreement was that i bring the Backplane to Taiwan into Qnaps HQ, as i plan to travel to Taipei anyway.
They are supposed to fix it and send it back to china mainland.
Costs for such thing ? Do not know until now.

Here is what happened yesterday, and that cracks me up.
I did install the new firmware (Yes, my TS-1079 Pro" is now EO Support and EO Firmware updates.... but there was one more......)

And as usual, before the FW flash, disk 10 was off, and Disk 9 as well, but Disk 9 got recognized every 10 minutes and dropped again after around 15 seconds.
Leds turned red at Disk 3,4,5,9,10...

But, after the install, which i did from a downloaded file, and i did choose not to do a reboot after install, i did reboot and here wo go, everything works, no disk drops....

So i am suspicious that there was at least one firmware that treated my 1079 pro as a 879 pro-....

I will reproduce that when i have time....
Answer to2:
You're supposed to put your existing drives into a new unit and everything shall work fine yes. So far to the concept.

Regards from China

[saliwok]

Thanks. I'll try one more firmware update and see what happens.
I was going to just ** it up and buy another unit, but it seems that every series of NAS's is having this issue, so probably not mechanical.

Let us know what you learn in Taiwan.