[resolved] clear "Disk Access History I/O" ERROR

[soosh]

@cameo Thanks for your solution. I was really stuck without it. In my case the drive was in excellent condition but the drive slot in my TS-431P was not. The NAS decided that the drive was defective after repeated disconnections. Moving the drive to a different slot was ineffective because the drive was effectively blacklisted. I was not interested in letting the NAS scan the drive for bad blocks because the drive had already been through 3 days of bad block testing connected directly to my PC.

In the process of getting the NAS working again, I bought a replacement drive. The failure of this drive to work in the bad drive slot is what revealed the real problem. Interestingly, the NAS did not block this drive when installed in the bad drive slot. There is probably some counter or heuristic that decides when to blacklist the drive.

Hopefully with a cleaning I will get the malfunctioning drive slot working again. At least my RAID array is working now I have a cold spare in case of real drive failure.

[Rawah1977]

All,

Great solutions posted here, but... I have a QNAP TS-451D2, firmware 5.0.0.2131. 4x WD 8T Red+ drives using RAID1. I have a similar issue as above, bay 4 drive (FEB 2022) failed SMART test ID197 "Current Pending Sector" and ID198 "Uncorrectable Sector Count" along with "Disk Access History (I/O) - ERROR. QNAP SMART Rapid Test failed "Read Failure".

I installed a brand new WD 8T Red+ drive in bay 4 and proceeded to REBUILD the Raid Drive.

This drive failed with a "cannot read"

I used WinSCP to log into the QNAP. I cannot for the life of me find /mnt/HDA_ROOT/.conf with the associated "pd_err_wwn_500xxxxx" lines.

What I see is - see the pics attached. If we go to WinSCP 03, there I see MNT/HDA_ROOT but no CONF file.

Anyone have any ideas where I might find an associated file with the "pd_err_wwn..."?

Anyone know how to search in WinSCP inside files for a word/line?

[Rawah1977]

All,

I have this issue and I am not finding this file /mnt/HDA_ROOT/.conf in my QNAP TS-451D2. Any ideas? Thanks.

[bbmike]

The file is hidden, so if you are using WinSCP and have navigated to the correct folder, can click where it says how many hidden files there are in the bottom right to show them, including this .conf

I had some major fun with all this, I was moving disks from a Synology to a shiny new QNAP TS-453E and had this Disk Access History error.
In my case, unbeknowns to me, the Synology wipe process (which I cba to wait to complete) puts a password on the HDD, so I had to use an old laptop with e-sata to connect these drives via an old external mount I had, reboot and remove the HDD password from the BIOS setup for each of the drives in turn.
But because I had already loaded them into the new NAS that .conf file was mashed up with all the locations the drives had been in when they had the error state.
Found best way was to elevate from a Putty session:

Code: Select all

sudo vim /mnt/HDA_ROOT/.conf

Can check how to use vi online, but I navigated to the rows in question then dd each line :wq and done.
All drives now clean and initialised fine.

[starnz]

All I am experiencing the same problem as this original post. I have a TVS-682 and had four 4TB HDD installed in RAID0. After running out of space on one of my groups, I backed up all my data and tried to replace one of my 4TB drives with a 12TB drive. I pulled the 4TB, inserted the 12TB, and after reading the brand new 12TB drive I get Disk Access History and Disk SMART Information errors

I'm assuming @cameo solution will work but I don't know how to do it. Is there a step by step guide somewhere that I can follow? I am capable just not familiar with the process.

[OneCD]

I have a TVS-682 and had four 4TB HDD installed in RAID0. After running out of space on one of my groups, I backed up all my data and tried to replace one of my 4TB drives with a 12TB drive. I pulled the 4TB, inserted the 12TB, and after reading the brand new 12TB drive I get Disk Access History and Disk SMART Information errors

Uh... if it really was RAID0, you just killed it by removing a drive. RAID0 cannot continue to operate with a removed drive member.

Time to restore from your backups.

[starnz]

Thats what I want to do, I backed everything up before starting. My problem is I need the 12TB drive to backup to but it is not accessible with the two errors. How can I access the 12TB drive?

[OneCD]

Thats what I want to do, I backed everything up before starting. My problem is I need the 12TB drive to backup to but it is not accessible with the two errors. How can I access the 12TB drive?

I'm unsure what the new 12TB drive has to do with your backup. Did you backup to the 12TB drive externally, then insert it into a NAS drive-bay in the hope you could read it again?

[starnz]

I backed up to an external drive. I'm trying to replace two 4TB drives that were a RAID0 group with a 12TB and 4 TB drive in RAID0. I got the original two 4TB drives backed up to the external. I then pulled one of those 2 4TB drives and replaced it with the 12TB drive. Now, I would like to restore the original 4TBx4TB RAID0 group info to the 12TB x4TB RAID0 group yet to be created.

Is what I'm trying to do possible? I read if I want to do RAID5 all the drives need to be the same size, but is that true for RAID0? that shouldn't have anything to do with the errors, right?

[OneCD]

I backed up to an external drive. I'm trying to replace two 4TB drives that were a RAID0 group with a 12TB and 4 TB drive in RAID0.

• If these 2 drives were part of a 4-drive RAID0, then it's not possible to replace any of the 4 drives. As soon as you remove a drive, the array is bust.

• If these 2 drives were a separate RAID0 from your other 2 drives, then you have the same problem. But at-least it's confined to 2 drives.

Best-thing (in both instances) is to kill the existing RAID0 before removing the drives, then insert your new drives and make a new array.

I then pulled one of those 2 4TB drives and replaced it with the 12TB drive.

Here is where it all went wrong.

Now, I would like to restore the original 4TBx4TB RAID0 group info to the 12TB x4TB RAID0 group yet to be created.

Can't be done in QTS. It doesn't support this.

You can backup and restore your shared folder data, but trying to apply a previous RAID config onto new drives is where you'll run into trouble.

Note: yes, it actually CAN be done, but you'll need to be an 'mdadm' guru, and know which parts of QTS to hack. Which will require you to be intimately familiar with QTS.

If you want to increase the size of your RAID0 array, you must destroy the existing array and create a new one, first ensuring all drives to be included in the array are the same capacity (see below).

I read if I want to do RAID5 all the drives need to be the same size, but is that true for RAID0?

Yes, they should be the same size. If you use different sizes, all drives are reduced in size to the capacity of the smallest drive (same limitation as RAID5).

You might investigate using a JBOD config instead to span each of your different sized drives. Then you'll have the full capacity available from each one. You'll be in the same precarious, dead-end position as you are now with RAID0 (as you'll have no redundancy), and you'll be putting even more of your data at-risk, but as long as you know the pitfalls, on your own head be-it.

that shouldn't have anything to do with the errors, right?

I agree. It sounds like you also have an issue with that 12TB drive. Did you confirm it's compatible with your NAS model?

[starnz]

The drive model is the same as my 4TB units, WD Red. I just assumed my errors could be cleared using the same procedure outlined by @cameo earlier in this post.

I'm not familiar with JBOD. I don't really care about redundancy as much a capacity. I have my dives backed up through HBS to externals so RAID array isn't important to me. Is there a way to get the full capacity of my 4 and 12TB drives using JBOD config?

[OneCD]

Is there a way to get the full capacity of my 4 and 12TB drives using JBOD config?

Yes, that's what JBOD is for.

[starnz]

I just pulled the 12TB drive and compared it to the 4TB I removed previously. There is a difference in connectors as the 4TB has a 6 or 8 pin connector that the 12TB doesn't. That must be the problem. Also, the 12TB is a 3.5"HDD while the 4TB is a 2.5". But they both have the same 3.5" form factor so I don't know what that means.

It looks like I need to return this 12TB drive and get 2 6TB Drives instead.

[giox969]

I had the same problem with QNAP QuTS Hero and zfs: a couple of zfs pools did not mount at boot for no reason. ZFS pools were ok, and disks had no errors. I was able to import the pools with zpool import from command line, scrub them, check status and everything was ok. But QuTS saved somewhere else that the pools had errors and refused to mount the pools.

The solution was:

• Edit /mnt/HDA_ROOT/.conf and remove all error lines containg

  Code: Select all

  pd_err_wwn_500xxxxxxxxxx541

  (as explained in previous posts here)

• Edit /mnt/HDA_ROOT/.config/qzpool.conf and set

  Code: Select all

  flag = 0x1

  inside the pools sections that QuTS is unable to mount (my QuTS had flag = 0x801 instead of 0x1)

• Reboot

Thank you @cameo for the help and @stefanonovello, @Qnap6470 for the hard work on testing and investigating.

[blub]

Hi

My 2c:

I had to roll back to QTS for this trick to work. In QTS5 there were no "pd_err" in the specified files - rolled back to QTS4 and there they were listed.