I have 4x 2TB SAMSUNG 850 PRO SSD's in a TurboNAS TVS-471, they have been starting to put out errors over the last week or two, so we purchased a set of new 4TB 860 Pro's. Friday (7/24) I went into Snapshot/Storage and did the "Replace disk one at a time" with Drive 4 as it gave off the most errors and figured i'd start there. Rebuild started just find and so I left it. Came back after about an hour and logged back into QTS (it had auto logged me out) to find that while the raid was still in read only mode, the interface was no longer showing it as in rebuild mode. This was a little worrying to me, but I figured it was just a bug. A few moments it later it gave me a pop-up saying their raid has failed and asked me if I wanted it to attempt to recover the raid. For some reason I believed this meant "attempt to continue rebuilding it. After reading up post-fact and realizing it was a system assuming the disk was orignally apart of the raid and not a new one trying to rebuild things went sideways. I the passwords no longer worked for admin and the portal no longer showed any assets aside from the login box. So I did what some recommended on the forum and shut down the NAS, unplugged the drives, turned it back on, then plugged the drives back in which allowed me to backup the passwd, shadow and group password files. I then restarted the NAS again to boot with the drives and was able to SSH in no problem. When I ran md_checker, it now shows this:
Code: Select all
Scanning system...
RAID metadata found!
UUID: d58e774e:e0d93d19:5d4ee57f:b2213d65
Level: raid5
Devices: 4
Name: md1
Chunk Size: 512K
md Version: 1.0
Creation Time: Apr 18 17:27:16 2016
Status: OFFLINE
===============================================================================================
Enclosure | Port | Block Dev Name | # | Status | Last Update Time | Events | Array State
===============================================================================================
NAS_HOST 1 /dev/sda3 0 Active Jul 23 20:18:48 2021 2660 AAA.
NAS_HOST 2 /dev/sdb3 1 Active Jul 23 20:18:48 2021 2660 AAA.
NAS_HOST 3 /dev/sdc3 2 Active Jul 23 20:18:48 2021 2660 AAA.
NAS_HOST 4 /dev/sdd3 3 Active Jul 23 18:02:28 2021 2362 AAAA
===============================================================================================
I've seen a number of people posting to run a command such as this: mdadm -CfR –assume-clean /dev/md1 -l 5 -n 4 /dev/sda3 /dev/sdb3 /dev/sdc3 /dev/sdd3
Not sure if this is the next step and want to check in as this raid array is very important to this business and I don't want to run the risk of borking it anymore than it has been (thanks to myself) if possible.