Problem with a QNap TS-212P :(

[LeKeiser]

I don't know your NAS but do you by chance have write caching running without a UPS? This is a no no. When power goes out it corrupts. Set write caching to write through. It is slower but much safer. Or use a UPS for your NAS. The same thing holds for Windows.

I haven't seen the option to set up write caching.
I bought a UPS few weeks ago but it doesn't change how files are being handled by the NAS. Or at least, I haven't seen any change.
But the duplicate files appeared before I set up the UPS. I bought the UPS after the power breakdown, as a matter of fact.

 

[coxhaus]

Once you have a corrupt HD there is no fix. Start over and run HD diagnostics. With RAID1 you can do 1 drive at a time if you don't have a backup. You will need to delete all the corrupted files which is not easy. But over time you will find them.

 

[LeKeiser]

Once you have a corrupt HD there is no fix. Start over and run HD diagnostics. With RAID1 you can do 1 drive at a time if you don't have a backup. You will need to delete all the corrupted files which is not easy. But over time you will find them.

Like I said, the disks are fine. I ran a thorough scan and the result was both disks are fine. SMART scans tell me the same thing.
The disks are not to blame. The system is

 

[coxhaus]

It is good the disks are mechanically fine but you still can have corrupted data. This where you normally go to a backup and start over.

PS
It is not the system but the user.

 

[dosborne]

I have other folders with accents, like Téléphone, Queensrÿche, ... and they kept the name,

The media server would be the most likely candidate if it only affects media files. But I'm still not convinved.

Is your QNAP firmware up to date?
TS-212P 4.3.3.1161 build 20200109 2020-01-09

 

[dosborne]

The disks are not to blame. The system is

You still have ZERO proof of that.

 

[dosborne]

SMART scans tell me the same thing.

What are the ACTUAL values for things like:

• SMART 5 - Reallocated_Sector_Count.
• SMART 187 - Reported_Uncorrectable_Errors.
• SMART 188 - Command_Timeout.
• SMART 197 - Current_Pending_Sector_Count.
• SMART 198 - Offline_Uncorrectable

https://www.computerworld.com/artic...that-actually-predict-hard-drive-failure.html

 

[LeKeiser]

What are the ACTUAL values for things like:

• SMART 5 - Reallocated_Sector_Count.
• SMART 187 - Reported_Uncorrectable_Errors.
• SMART 188 - Command_Timeout.
• SMART 197 - Current_Pending_Sector_Count.
• SMART 198 - Offline_Uncorrectable

https://www.computerworld.com/artic...that-actually-predict-hard-drive-failure.html

Disk 1 :

• SMART 5 - 100
• SMART 187 - 100
• SMART 188 - 100
• SMART 197 - 100
• SMART 198 - 100

Disk 2 :

• SMART 5 - 100
• SMART 187 - 100
• SMART 188 - 100
• SMART 197 - 100
• SMART 198 - 100

So as you can see, SMART results are good, disks are fine

 

[L&LD]

SMART values only try to describe physical properties of the HDD.

Corruption is logical, as people keep trying to tell you above.

 

[dosborne]

So as you can see, SMART results are good, disks are fine

All value are 100? Looks fishy. I'll have to take a look at mine tomorrow. What about run time?

 

[LeKeiser]

SMART values only try to describe physical properties of the HDD.

Corruption is logical, as people keep trying to tell you above.

What do you mean, logical... Why only some directories and not all of those having accents in their names? Why did those names changed when I wasn't using those directories? Why the duplicates after a power shutdown, when all my other NAS and my server running Debian never ever ever had duplicates after a power shutdown. Never seen that also with all the servers I monitor at work, and trust me, I have a few...
You say logical? Read all the posts above. People seem to think it's a hard disk failure, hence all the physical tests and the SMART test.
I'm sorry, don't mean to be harsh, but I know it happened because of the QNAP OS, the way it's handling the Raid part. And from what I see, it's not good, certainly not the best.

 

[LeKeiser]

All value are 100? Looks fishy. I'll have to take a look at mine tomorrow. What about run time?

I'm not sure I can install smartmontools on the QNAP, it would give me a more thorough result. But those I gave you are the ones QNAP showed me.
The NAS has been updated last week, so the uptime is short, say a week at best.

 

[LeKeiser]

The thing I really want to undertand with QNAP, is how their Raid application works.
Raid1 should duplicate the files on the second drive. Nothing more. It shouldn't duplicate or triplicate or quadricate the files on both drive. It's not what it's made for.
How come only some files? Why not all the files?? Doesn't make sense.

Raid1 shouldn't change the names of the files. Just because I lost the network, doesn't explain why when I got it back, some directories and some file names got their names changed. Doesn't make sense too. This part looks like it's more OS related but that would mean something is missing in the QNAP OS. Or badly configured.

That's why I'm not happy with QNAP and seriously thinking to change this NAS.

 

[L&LD]

You're looking for blame where there seem to be none?

If your issue was a QNAP issue, I think I would have run into it in the last decade from at least one of my own QNAP NAS's or my customers.

When was the last time you did a full reset of this device? Without stripping it of its tools to manage itself?

It seems to me your unhappiness is self-inflicted, honestly.

Btw, an HDD can be working perfectly at the hardware level, but at the logical/software level, something may be amiss.

You seem to only want to acknowledge one of the above situations.

 

[LeKeiser]

You're looking for blame where there seem to be none?

What are you saying? I have duplicates/triplicates/... files and you're saying there is nothing to blame?
The names of the directories and the files change for no reason and everything is peachy and I shouldn't worry??
Come on... Please!

If your issue was a QNAP issue, I think I would have run into it in the last decade from at least one of my own QNAP NAS's or my customers.

Do you use the same model I do? Same version of OS?
Oh wait... You're going to tell me that you work in the business, you have decades of high level IT and you, out of all the people, know better...
Funny, I keep on running into those people a lot these days.

When was the last time you did a full reset of this device? Without stripping it of its tools to manage itself?

What do you mean, a full reset of the device. Want me to break the Raid1 and start over? It't not my way of fixing things... And if it's yours, well I won't follow your advice. Sorry.

It seems to me your unhappiness is self-inflicted, honestly.

Again, duplicates/triplicates/... , random change of names, ...
and I'm self-inflicting this?
Come on. Please!

Btw, an HDD can be working perfectly at the hardware level, but at the logical/software level, something may be amiss.

And that's what I've been saying since the begining.
I know the disks are fine. I've tested them many times.. For what they are worth, I've checked the smart results and they are fine.
It's a problem with the way QNAP is handling the Raid part, or a really weird problem with the ext4 partition. Something I have never seen in my 25 years working on Linux (yes I started with Slackware back in 1994).
I haven't gone deeper into the QNAP OS, didn't have time and was too pissed off when I got those name change, but I might have a look, out of curiosity. Or not.

You seem to only want to acknowledge one of the above situations.

I don't acknowledge anything.
I would like to know if someone had the same problem before.
Some say it's a hardware problem.
Some say it's a logical problem. I guess they're saying it's OS related. And I think they are right. But there must me a way to prevent that. And that's what I'm looking for.
Or I'm just gonna change this NAS and buy another brand.
Never had those problems with my WD or my DLink. And they went down the same way the QNAP did.

But I might be self-inflecting some unhappiness

 

[L&LD]

@LeKeiser, sorry, but you are assuming a lot of things from my brief posts, while also ignoring any suggestions to help you.

I wish you luck.

 

[LeKeiser]

@LeKeiser, sorry, but you are assuming a lot of things from my brief posts, while also ignoring any suggestions to help you.

I wish you luck.

Thanx.
I don't think I'm assuming anything from your post. I've tried to reply to your post the most honest way and as best as I could. Didn't mean to offend you or piss you off.

Thing is, I happen to know also a bit about Raid stuff, Linux stuff, NAS stuff, Windows stuff,... But I don't consider myself an expert. So I came here to get some feedback from you guys. As long as they are valid, or interesting, or... can't find the words, at work and busy but I want to respond fast.

Breaking the Raid1 partition is a no go for me. I don't see why it has something to do with the partition. Again, I think it's something wrong with the ext4 partition, or the way it's being handled. Some missing configuration, or something wrong in the configuration. Might be some locales indeed. Will have to investigate. When I have time.

 

[L&LD]

I'm not pissed off.

But it doesn't seem like anybody can get their points across to you?

Please, re-read the thread here. I'm certainly no expert either. Just trying to help as best I can.

 

[LeKeiser]

I'm not pissed off.

But it doesn't seem like anybody can get their points across to you?

Please, re-read the thread here. I'm certainly no expert either. Just trying to help as best I can.

but what points?
Some are saying it's a hardware problem. Bad disks. But if it was a problem with a disk, then the scans I have done should have reported something wrong. Bad sectors, corrupt tracks, ... And everything is fine with the disks. With both of them.
So it's not a hardware problem.

And since the Raid is a software Raid, it might have something to do with the configuration. And that is what I believe people are thinking when they say "it's a logical" problem.
And I would love to know more about that.
That or the ext4 configuration.
Cuz I am sure my problems come from one of those 2 things. Raid configuration or ext4. Has to be.
But then, that's a QNAP issue...

 

[dosborne]

Again, duplicates/triplicates/... , random change of names, ..

To reiterate, this is normal recovery for a failure in raid mirroring. This is a raid function, you can "blame" QNAP if you wish, but it is basic Linux code, nothing to do with them really.

I'll also ask again what HDD Manufacturer and model you are running along with the approximate age of the drives.

The S.M.A.R.T. values of 100 that you quoted are not the actual values, you need to look at the "raw value" column
https://www.z-a-recovery.com/manual/smart.aspx

The NAS has been updated last week, so the uptime is short, say a week at best.

It isn't the QNAP uptime I was asking about, it is the runtime of the HDDs.
SMART 9 power-on_hours.

How about 1 raw_read_error_rate and 7 seek_error_rate?

But if it was a problem with a disk, then the scans I have done should have reported something wrong. Bad sectors, corrupt tracks, ... And everything is fine with the disks.

No, not necessarily. If the HDD self-corrected, then your tests would NOT show this. It would be buried in the SMART data.

I have both a TS-563 and a TS-231P (the later is pretty similar to yours). In all fairness, this is primarily a router forum. Have you tried posting on the QNAP community forum https://forum.qnap.com/ or entering a help ticket with QNAP directly? (I strongly suspect that they will tell you the same things we are though).