Raid 6 with Error, 2 discs out, volume not active, can't recover after inserting new drives

[roly128]

Hi all,

I'm at a loss here. TS-431P with 4x2GB disks in RAID 6 configuration. FW 4.4.3.1354 (Old but I did not dare updating while having a disk error).
First disk 2 issued a warning. I ordered a spare disk to replace it. While waiting for the new disk to arrive, disk 3 went into error taking the volume with it deactivating file system access.
Ordered a second spare disk and replaced both defective discs (I thought that should be possible with RAID6). But the volume is still not active and won't recover. Option to recover is selectable but aborts with an error.

Can sombody pat my hand and tell me everythings gona work out? Really don't know where to go from here. What should I do?

(I have basic knowledge how to open a command line and log in via SSH, but I'm no CS, if that helps)

Thanks!

[dolbyman]

2 disks should kept the raid active...if it did gave out, more disks might have issues..hope you have backups

do an

Code: Select all

md_checker

via ssh to check the status

[roly128]

Thats the strange part, disks 1 and 4 show no degradation. Might the NAS it self be dying?

md_checker outputs the following:

Code: Select all

md_checker 

Welcome to MD superblock checker (v2.0) - have a nice day~

Scanning system...


RAID metadata found!
UUID:		48d4cc0c:3e681c64:fd42b3e0:29db8b39
Level:		raid6
Devices:	4
Name:		md1
Chunk Size:	512K
md Version:	1.0
Creation Time:	Sep 19 20:12:51 2018
Status:		OFFLINE
===============================================================================================
 Enclosure | Port | Block Dev Name | # | Status |   Last Update Time   | Events | Array State
===============================================================================================
 NAS_HOST       1        /dev/sda3   0   Active   Nov 15 07:31:05 2021   111373   A..A                     
 ----------------------------------  1  Missing   -------------------------------------------
 ----------------------------------  2  Missing   -------------------------------------------
 NAS_HOST       4        /dev/sdd3   3   Active   Nov 15 07:34:32 2021   111374   ...A                     
===============================================================================================

All disk LEDs are green and are reported to be healthy. But the storage pool reports Error (Not active)

Edit: Please note that all 4 disks are installed and running. 1 and 4 are the original ones 2 and 3 are new ones. All show healthy.

[dolbyman]

try a

Code: Select all

mdadm --assemble --scan 

[roly128]

Thanks for your help!

mdadm --assemble --scan yielded:

Code: Select all

mdadm --assemble --scan
mdadm: scan_assemble: failed to get exclusive lock on mapfile
mdadm: No arrays found in config file

same when I try with sudo

[dolbyman]

I would open a ticket with qnap then

[roly128]

hmm I see, looks quite hopeless then... Thanks anyway for your effort in helping out, much appreciated!

[dolbyman]

might be something simple, not sure what it could be though

[roly128]

I have issued a support ticket, will post the solution if I get one. Thanks again!

[roly128]

Supporter has asked me to issue the following, in order to bring the storage pool back online:

Code: Select all

mdadm -AfR /dev/md1 /dev/sd[ad3]

Once it's back online it should be followed by this:

Code: Select all

/etc/init.d/init_lvm.sh

Unfortunately the first command was to no avail for me. I'll document it notheless, as it might help someone somewhere sometime... This was my systems reaction:

Code: Select all

[/] # mdadm -AfR /dev/md1 /dev/sd[ad3]
mdadm: Cannot assemble mbr metadata on /dev/sda
mdadm: /dev/sda has no superblock - assembly aborted

[dolbyman]

sounds like one of the remaining disks (sda) has issues too

[roly128]

Hi all and especially dolbyman,

The issue has been resolved after QNAP support remotely restored the volume. No further explanation as to how they fixed it.

However disk 4 now also reports issues, and I'll backup the whole volume before going any further. Support strongly suggested to wipe the entire volume and start from scratch once I've backed up all data.

Quite strange that 3 out of 4 disks presented issues almost at the same time. They were produced and purchased at the same time too, so it might be a production issue? All disks were from the type Seagate Barracuda ST2000DL003, 2000GB, Date: 12343, Site: WU.

Thanks again for the help, much appreciated!

[Toxic17]

glad you got it solved. do you run SMART tests on your drives as a schedule? I tend to do the rapid test weekly and the complete test monthly. better to catch a predictive fail than to have disks fail without any warnings.