Error: Invalid argument during seek for write on /dev/sdg

TrueNAS General
, ,
1

I have searched every way I know how for a fix for this but I cant find squat.

I have a 6 disk, 1 vdev pool. Currently, it uses mixed storage because I had a few of my SATA SSDs (5 Samsung 870 EVO, 1 Crucial MX500) fail spectacularly. I swapped in some SAS spinning rust (cheetah 15k) to hold things together while my new SSDs got here. Well, my new SSDs got here yesterday (Crucial MX500s running FW: M3CR046) and when I try to swap one in to replace the spinning rust and get the pool back to a somewhat homogeneous makeup, I get the error in the title. I am relatively new to TrueNas so I am not sure how to diagnose errors like this as the shell seems to be missing my staple tools like gparted and fdisk.

My big issue is that I have this pool exported as NFS for a proxmox storage node so Every time I do something that requires a reboot, its an hour long ordeal. ( Thank you, project Zomboid, for making your dedicated server take 20 minutes to terminate. )

Any pointers are appreciated and I am willing to provide diags if someone can point me to the correct manual to read.

Edit: Where are my manners? God, you would think I am new to the idea of forums.
NAS hardware:
Dell Optiplex T420
2x Xeon E5-2470 v2
48GB RAM
PERC H710 flashed in IT mode

Software: TrueNas Dragonfish. I stay updated.

2

Well, now this is dumb. And maybe something the devs should be made aware of.

Somehow, by some horrible miracle, my box enumerated two disks as /dev/sdg/.

I have no idea how, but I caught this after looking over the pool once more. So I shut down the box, disconnected all disks, waited for the thing to power drain, and then spun it back up and the disk enumerated correctly. I am leaning HBA glitch but who knows.

I have been working with Linux for nearly 20 years now and I have never once seen that so anyone’s guess is valid.

Also, go figure I sort this out as soon as I post to the forums. C’est la vie.

1 Like