Page 1 of 1

[Disk]There is a data transfer error on sector ... of the slot ... disk

Posted: Mon Jul 23, 2018 1:54 pm
by Gunthor
This morning after powering up the NAS I received the following E-Mail:
"This is a system event notification sent from NAS.

Event Level: ERROR
Date: 23.07.2018 07:18
User: SYSTEM
Event: [Disk] There is a data transfer error on sector 88824064 of the slot 2 disk!"
The LED of slot 2 shows the red light :!:

In the "Overview" the Storage manager shows the error as well:
Image

Whereas in the "Disk" section slot 2 is getting displayed as functional and healthy:
Image

Within the Asustor Knowledgebase...
https://support.asustor.com/index.php?/ ... es-it-mean
... I found the following advice:
For [...] transfer [...] error, users are advised to have some more clarification to identify if the problem is caused by hard drives or NAS hardware. Users can swap the error drive with another normal drive, and check if the log will report same slot error.For example, Users found that the 1st slot reports transfer error, then he/she can shutdown the NAS, switch slot 1 & 2, power on the NAS again, and check if the error moves to slot 2.
If the error moves to slot2, that means this drive might have some bad sector and users might want to replace it. If the error stays with slot1, users might have to contact with Asustor support for further investigation.
Questions:
Is this advice Raid independent :?:
Will this also work with Raid 5 on a 4 slot device :?:

Re: [Disk]There is a data transfer error on sector ... of the slot ... disk

Posted: Mon Jul 23, 2018 4:29 pm
by orion
Volume-1 is still in healthy state. The transferring error might be a recoverable error in volume-1. Or the transferring error might be a non-recoverable error in system OS partition which is mirrored to all of your disks. It's better to check /proc/mdstat.

Re: [Disk]There is a data transfer error on sector ... of the slot ... disk

Posted: Mon Jul 23, 2018 5:52 pm
by Gunthor
Thank you @Orion for your hint. However, I can't say from looking at mdstat what to do next:

Code: Select all

Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4] 
md1 : active raid5 sda4[0] sdd4[4] sdc4[3] sdb4[2]
      8777425344 blocks super 1.2 level 5, 64k chunk, algorithm 2 [4/4] [UUUU]
      
md126 : active raid1 sda3[0] sdd3[6] sdc3[5] sdb3[4]
      2096064 blocks super 1.2 [4/4] [UUUU]
      
md0 : active raid1 sda2[0] sdd2[6] sdc2[5] sdb2[4]
      2096064 blocks super 1.2 [4/4] [UUUU]
      
unused devices: <none>
No slot is marked with (F)! So can I lean back trusting that everything is OK? Further suggestions are highly appreciated.. :)

Re: [Disk]There is a data transfer error on sector ... of the slot ... disk

Posted: Tue Jul 24, 2018 10:00 am
by orion
Then, the error should be a recoverable error. Yes, I believe your system is still in secure / protected state.
If I were you, I'll check "dmesg" to make sure what's the error.

Re: [Disk]There is a data transfer error on sector ... of the slot ... disk

Posted: Sun Aug 05, 2018 9:06 pm
by Gunthor
OK - on 24/07/2018 the bad block scan finished and I got an error:
Image
Then I switched the HDDs: slot 1 <-> slot 2 and repeated the bad block scan and can confirm that the Asustor advice (see first post) is Raid independent: The Raid is still functioning and now the bad block error is no longer on slot 2 but on slot 1. This confirms that it is not caused by the NAS hardware. :)
Also, the red LED light next to the slot does not show red any longer. :)

As for the "dmesg" command the -e option is not sustained: No human-readable time format gets displayed and I couldn't find the cause of the error. :(

Asustor suggests to replace the disk (see first post). However, proc/mdstat continues to not show an (F) for any slot. :)

Questions:
Is my assumption correct, that even though there are bad blocks the disk (apart from the bad HDD blocks) is working within the Raid :?:
So only if I am not willing to risk that the same information (that was stored on the bad blocks) gets lost on one of the other disks (and then would be gone permanently - apart from existing backups) I should replace the disk that shows the bad block - correct... :?:

Re: [Disk]There is a data transfer error on sector ... of the slot ... disk

Posted: Mon Aug 06, 2018 9:40 am
by Antonios
Man, Murphy's laws you know...

I would get rid of that HDD even if it was part of a RAID 6. Maybe I love my data too much.

Just my 2c.

Re: [Disk]There is a data transfer error on sector ... of the slot ... disk

Posted: Mon Aug 06, 2018 11:04 am
by orion
Looks like that the bad block is not located onto RAID regions. I wonder how system can detect the bad block error without scanning. Or you setup a schedule bad-block-scan task? In that case, system should access non-RAID regions.
Not sure if this bad block is important or not. If it's partition table, that is a fatal error. Anyway, you'd better to replace the failed disk.

Re: [Disk]There is a data transfer error on sector ... of the slot ... disk

Posted: Mon Aug 06, 2018 6:30 pm
by Gunthor
Thank you for your response @Antonius and @Orion,
especially as I have a spare disk waiting to be used I will do just that and will replace the faulty one tonight. ;)