access to Readynas admin page impossible, one disk looks dead

Hi Vincefr34 

This does not look like a filesystem side issue. Hence you running the "btrfs rescue zero-log" command would not fix this. That is also what you reported in the PM to me, as you tried this on your own accord. It fails to actually run the command because the filesystem complains about a missing device.

root@SERVEUR:~# btrfs rescue zero-log /dev/md126
warning, device 1 is missing

Looking at the dmesg log it can be seen that md127 tries to start with sda3 and sdb3 but that fails due to not enough partitions to start the raid.

[Sat Apr 23 17:55:52 2022] md/raid:md127: not enough operational devices (2/4 failed)
[Sat Apr 23 17:55:52 2022] RAID conf printout:
[Sat Apr 23 17:55:52 2022] --- level:5 rd:4 wd:2
[Sat Apr 23 17:55:52 2022] disk 0, o:1, dev:sda3
[Sat Apr 23 17:55:52 2022] disk 2, o:1, dev:sdb3
[Sat Apr 23 17:55:52 2022] md/raid:md127: failed to run raid set.

It reports that md127 only see 2/4 partitions, but the RAID5 would require 3/4 partitions. Due to different sized disks, you have multiple raid instanced which the filesystem "glues" together. It appears that you have 3 different raid instances for the data. One of these raids can't start and that causes the entire data volume to be inaccessible.

You have enough disks in the box to start these raids and the disks are healthy. I very much suspect the issue is that the partition table on sdc (Disk 1) is messed up. It is missing partitions... definitely missing sdc3 and possibly sdc4 as well. But the lack of sdc3 is the reason that the raid fails to assemble.

8 16 7814026584 sdb
8 17 4194304 sdb1
8 18 524288 sdb2
8 19 3902297912 sdb3
8 20 976754027 sdb4
8 21 2930255992 sdb5
8 32 7814026584 sdc
8 37 2930255992 sdc5

We have seen before that partitions goes missing, for unknown reasons. I cannot see any health issues on the remaining disks and the logs don't immediately indicate why this happened. The solution would be to copy the correct partition layout (likely from disk 2 - sdb), to disk 1 (sdc). That should allow the raids to assemble properly.

However, one should look at the md-superblocks to ensure this is done correctly. Netgear should assist with this, realistically. I agree with StephenB on that. They used to offer some data recovery service, and I think they still do. Not sure how you would go about making a ticket these days. Perhaps StephenB knows (or maybe we need to tag the forums mods?)

Cheers