It was inevitable…
Not long after setting up our TrueNAS server I logged into the web interface and noticed that one of the drives showed an error. I almost expected this as one of my 8TB Seagate Ironwolf drives is a slightly different model, and was older than the rest. When I looked up the warranty, it had apparently expired in 2020 (leaving me wondering when I bought the drive because I didn’t think it was much older than 4 years).
I was a bit concerned as I finally managed to transfer most of our media collection to the server, and already I had to replace a drive (only a couple of months later). 8TB Seagate Ironwolf drives go on and off sale frequently. As of this post the least expensive I’ve seen a new one is $219 CDN, but they generally hover around $249 CDN on sale. Just before Christmas the price shot up to $279 CDN on sale making me question the practices of Seagate (as the price at most places seems to be controlled by Seagate themselves). Were it not for the shady SMR stuff Western Digital did with their Red drives awhile back I might have looked at adding a WD Red 8TB to the mix.
ZFS Scrub?
I still have a lot to learn about TrueNAS. One of the impressions I got was that doing a “ZFS Scrub” within the web interface sometimes fixed issues with drives. After learning about the error, I tried running a ZFS scrub a couple of times. The drive still showed errors after the scrub.
Inventorying drives
I took a quick inventory of the drives, writing down the model numbers, and quickly checking the warranty status. All the drives were under warranty, except the drive that had the errors. Early on, non-Pro Ironwolf drives did not have any “Rescue Service” warranty, but the last two drives I purchased apparently do. The errors made logical sense as they were on the oldest drive in the bunch.
Merry Christmas, Happy New Drive
Santa was very good to me, and got me a new 8TB Seagate Ironwolf for Christmas. After a month or so of worrying about the 8TB of data stored on the NAS, that worry was a bit lessened in my mind. Now that I had a replacement drive I needed to investigate what I needed to do on TrueNAS to prepare the old drive for replacement. As I usually do for server stuff, I looked to a Lawrence Systems video for a tutorial on How to Replace A Failed Drive in TrueNAS Core or Scale. Then I hopped over to our TrueNAS server and … wait … what’s this??? No errors on any of the drives in the array.
Perhaps RAM ???
I’m not sure how the errors seem to have fixed themselves over the past month or so, but I’m starting to think that perhaps one of the reasons for the errors was the fact that I’m using 32GB of non-ECC (Error Correcting Code) memory. I don’t know, but the system I’ve built won’t take ECC RAM. I do have access to other hardware that accepts ECC RAM, but after the effort I went to build in the Fractal Define R5 Case I don’t really want to mess with it too much. Those fan cables are pretty stretched as is.
Whatever the case I now need to test this new drive and figure out what to do with it. I’m going to continue to monitor the NAS as I expect the errors might show again as I transfer more data. I have about 100GB of data to transfer this morning. I’ll keep posting about the NAS from time to time, but this is an oddity.