Hi folks. We’re running Scale 23.10 (can’t update at this stage because of an SMB extended-attributes bug in 24.10+ that breaks our backups) on two machines with 84 HDDs and 6 SSDs. Our layout is 11 vdevs of 7 drives in a RAID-Z2 with 7 spares, 3 log SSDs and 3 metadata SSDs.
3 months ago, we had a drive failure in one machine and it was replaced. While doing some maintenance, I just noticed something strange - the pool spec shows 8 spares, not 7. zpool status from CLI shows the exact same. I checked the serial numbers of the drives, and sure enough, it’s the replacement drive - it’s showing in both the vdev, where it should be, and in the Spares list. All of the vdevs show 7 drives.
I don’t know if this is harmful but it’s certainly confusing. What’s weirder is that these two machines are multi-site replication partners; the one with the replaced drive is the target, but I noticed the same thing on the source, it’s also showing 8 spare drives. We had a similar disk failure and replacement 5 months ago, but the extra spare drive in this instance is not the replacement.
Anyone have any suggestions for whether we need to do anything here? Like I said, we cannot currently upgrade these machines from 23.10; we are replacing our backup software and will eventually be able to upgrade, but not right now.
Cheers,
Rob