Deleted an account used for TimeMachine and my RN is now in ReadOnly Mode...

[Zecaf]

Hello all...

New fun one: deleted an account in the WebUI (Firmware 6.10.1). I use accounts only to manage different TimeMachine backups. The Delete seems to be ok. Then my Mac pop an error problem on the volume. Log says : Volume: The volume data encountered an error and was made read-only. It is recommended to backup your data.

I go check my TimeMachine backups : fun fact, the account in gone ideed BUT the backup of another account shows no data (It was fine just before), like it was deleted at the same time. Hardware seems fine, FS seems going to **bleep** for no reason. Since I already lost the whole thing last year, I'm managing backup right now... Any idea about the problem? Any command to run to check the integrity of the file system? OFC I'll format the whole thing post-backups but this bug is crazy... :x 

[Zecaf]

I'm surprised (and sad, let's face it) that no one has the smallest information or tips to share about this problem. 😕

[Marc_V]

@Zecaf

Can you send me the full logs. There might be something else that caused the NAS to be on Read Only mode.

Regards

[Zecaf]

Thanks for your answer. I sent you a private message with the link to the logs.

I hope you'll dig something useful. 🙂

[Zecaf]

Still no idea?

I tried a scrub, nothing is fixed.

dmesg shows :

[  105.003830] BTRFS error (device md127): parent transid verify failed on 718766080 wanted 271645 found 322138
[  105.011617] BTRFS error (device md127): parent transid verify failed on 718766080 wanted 271645 found 322138
[  105.011628] BTRFS warning (device md127): Skipping commit of aborted transaction.
[  105.011637] BTRFS: error (device md127) in cleanup_transaction:1864: errno=-5 IO failure
[  105.011641] BTRFS info (device md127): forced readonly
[  105.011645] BTRFS: error (device md127) in btrfs_drop_snapshot:9412: errno=-5 IO failure
[  105.011749] BTRFS info (device md127): delayed_refs has NO entry

And btrfs check --force /dev/md127

WARNING: filesystem mounted, continuing because of --force
Checking filesystem on /dev/md127
UUID: 9d4171ad-fe40-45a0-9013-12e3db1a26fa
checking extents
checking free space cache
checking fs roots
checking csums
checking root refs
checking quota groups
qgroup-verify.c:538: account_all_refs: BUG_ON `ref->num_bytes != num_bytes` triggered, value 1
btrfs[0x42ad55]
btrfs[0x42ad74]
btrfs[0x42ae39]
btrfs(qgroup_verify_all+0x448)[0x42c0bb]
btrfs(cmd_check+0x25c8)[0x44e842]
btrfs(main+0x13d)[0x40c647]
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5)[0x7fd2ab413b45]
btrfs[0x40c67b]
Aborted (core dumped)

I had another error before my last reboot, related to qgroups too...

My last question if someone is still around: what is the safest way to start fresh? Factory default? I guess destroying the volume would not be enough to be sure all is back to "normal"? Thanks if you have informed inputs. 🙂

[StephenB]

I'd do the factory default - it doesn't take much longer, and gives you a fully clean OS partition.

[Zecaf]

Yeah I did that and that was enough to kill another HD from 2015, 6 month after the warranty expired. Thanks Seagate (sector errors). I did an extanded drive check and ordered a new HD. Dusted the damn thing and the next time I have a problem, the whole thing hit the trash and I upgrade to a Syno. I'm tired of this. Really (and I've been defending ReadyNAS then Netgear for almost 20 years...). (To be clear, this is my 3rd major problem in less than 4 years with no support or remotly usefull help, and icing on the cake, this whole support forum framework is horrible, we're in 2019 and stuff like Discourse are FREE...)  End of rant. Sorry. :x

[Sandshark]

The cause of the read-only was probably the failing drive.  The process of recovering free space from the deleted account may or may not have accelerated that.  When certain types of access problems occur, it goes to read-only to try and give you an opportunity to backup your data before it fails completely (which writes could accelerate).  How you figure a failing drive is Netgear's fault, I don't get.

While BTRFS has a lot of safeguards, once a file system is "broken", recovery isn't always possible, even if the root cause is fixed.  I had a volume in an EDA500 go read-only because of a bad connection of the eSATA cable.  Though I fixed the cable issue, I could not make the volume read-write -- I had to destroy and re-create.  It had already benn corrupted too much/  Still, had I not already had a backup, I was presented with an opportunity to do so instead of just a dead volume.

I just recently had two WD Gold Enterprise drives fail, one 3 months past warrant and the other 5.  I hjad problems with Seagates long ago and have avoided them for some time.  Though now also a part of WD, my Hitachi/HGST drives seem to have the best longevity.

[st_pusc]

Same problem here, last tuesday I deleted an account used for Time Machine only (1TB reseved) and then I couldn't connect via AFP, I managed to connect with samba and then I found that the filesystem was read-only... At least I could copy almost everything, bit by bit because if I tried to copy a big amount of data the NAS hang and the only way to restart it is unplug the power...

My firmware was 6.9.2, there are 4 HD WD-Red 2TB, in raid5 configuration. It has been running since 2015

# mdadm -D /dev/md127
/dev/md127:
Version : 1.2
Creation Time : Sat May 9 10:28:11 2015
Raid Level : raid5
Array Size : 5845988352 (5575.17 GiB 5986.29 GB)
Used Dev Size : 1948662784 (1858.39 GiB 1995.43 GB)
Raid Devices : 4
Total Devices : 4
Persistence : Superblock is persistent

I tried everything, frome the web interface and ssh (test data, scrub, delete cache, repair...). repair always abort.

Updated to 6.10.2 hoping that repair will work but same situation...

The hard disks look OK, I ran smartctl -a -d sat in each of them, no errors and there was a extend test maybe triggered by one of the actions I took from the web interface.

SMART Self-test log structure revision number 1
Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
# 1 Extended offline Completed without error 00% 30866 -

So the hardware doesn't seem the problem.

I have been using ReadyNAS for more than 10 years, I have 5 of them now... I won't delete any time machine back up in the future and I will use my owns linux system witn XFS for storage....

ST