I've recently moved and still unpacking. Turns out my BIOS battery is dead. Also turns out that an idiot decided that its best placed under the heatsinks of the motherboard. Its a complete removal of the motherboard to get to it for a replacement. So, I have to set up the bios each time I down the server.

On top of this, one of my array's drives has gone offline and a second is reporting reallocated sectors, slowing increasing over the last few months.

I can still access the offline drive fine. I'm backing it up at the moment. The server was unresponsive this morning, whilst doing the last backup. I think the C-state power setting in the bios was the culprit (owing to the bios resetting).

However, I've come across this message below in the logs and I'm not quite sure what to do with it. Yes, I'm aware there are instructions, which may not be suitable in every case, I would rather ask.

Apr  2 20:16:24 Nexus kernel: mdcmd (31): set md_num_stripes 1280
Apr  2 20:16:24 Nexus kernel: mdcmd (32): set md_queue_limit 80
Apr  2 20:16:24 Nexus kernel: mdcmd (33): set md_sync_limit 5
Apr  2 20:16:24 Nexus kernel: mdcmd (34): set md_write_method
Apr  2 20:16:24 Nexus kernel: mdcmd (35): start STOPPED
Apr  2 20:16:24 Nexus kernel: unraid: allocating 36230K for 1280 stripes (7 disks)
Apr  2 20:16:24 Nexus kernel: md1p1: running, size: 17578328012 blocks
Apr  2 20:16:24 Nexus kernel: md2p1: running, size: 5860522532 blocks
Apr  2 20:16:24 Nexus kernel: md3p1: running, size: 13672382412 blocks
Apr  2 20:16:24 Nexus kernel: md4p1: running, size: 13672382412 blocks
Apr  2 20:16:24 Nexus kernel: md5p1: running, size: 17578328012 blocks
Apr  2 20:16:24 Nexus emhttpd: shcmd (75): udevadm settle
Apr  2 20:16:24 Nexus emhttpd: shcmd (81): xfs_admin -U generate /dev/md4p1
Apr  2 20:16:24 Nexus root: ERROR: The filesystem has valuable metadata changes in a log which needs to
Apr  2 20:16:24 Nexus root: be replayed.  Mount the filesystem to replay the log, and unmount it before
Apr  2 20:16:24 Nexus root: re-running xfs_admin.  If the filesystem is a snapshot of a mounted filesystem,
Apr  2 20:16:24 Nexus root: you may need to give mount the nouuid option.  If you are unable to mount
Apr  2 20:16:24 Nexus root: the filesystem, then use the xfs_repair -L option to destroy the log and
Apr  2 20:16:24 Nexus root: attempt a repair.  Note that destroying the log may cause corruption --

I've received two Ironwolf 24TB drives today. Which are larger than my current drives. My intention is to:

• preclear both drives, whilst I re-add the offline drive (I think it could be a loose cable issue that seems to be okay). Hopefully, rebuilding this drive will be okay.

• Then replace the 20TB parity drive with a new 24TB drive. Use the 20TB to replace the offline drive.

• Add a new 24TB to replace the drive with re-allocated sector issues.

  Because of the size of the replacement drives, I cannot think of another way to do this. Unfortunately, drives are in such short supply in Australia, I was lucky to get these.

  However, I really need to know what to do with the file system xfs_admin issue? I don't want to start issuing these commands if there is a risk of further damage, especially when my drives are looking unreliable. Help appreciated.

• Diagnostics attached. I recognise its Easter, but help always appreciated.

nexus-diagnostics-20260402-2114.zip

Disk4 is mounting, so you can ignore that for now, but when you can, check filesystem on that disk.

Thanks JorgeB!