First aid Error

I've had a few drives where it said there were errors that were repaired, but it said the same errors were there when I ran first aid again.

Can't be sure what's going on, but it might make sense to back up your drive just in case.

I've seen multiple posts on these forums with this complaint and no resolution. The APFS file system is new and I think Apple is still working out the bugs. Disk Utility has never been great at repairing file systems in the past anyway. Unfortunately Apple has not released the necessary APFS documentation so that third parties can update their file system repair utilities to work with APFS volumes.

If you are not having any issues, then this could just be a Disk Utility bug, but I would definitely follow @y_p_w's advice and have good verified working backups just in case. Backups are even more important when using SSDs since an SSD can fail without any warning and data recovery can be difficult to impossible.

Your other option is to make a backup and erase the physical drive and reinstall macOS. Do not erase the "Macintosh HD" volume as that may not be sufficient to get rid of the errors (if they really exist). When in Recovery Mode and within Disk Utility you need to click on "View" and select "Show all devices" so that the physical drive shows up in the left pane of Disk Utility.

HWTech wrote:

If you are not having any issues, then this could just be a Disk Utility bug, but I would definitely follow @y_p_w's advice and have good verified working backups just in case. Backups are even more important when using SSDs since an SSD can fail without any warning and data recovery can be difficult to impossible.

As someone who has worked in the SSD industry, in general SSDs are more reliable than hard drives because of little chance of mechanical failure. Flash storage (especially tri-level NAND is inherently unreliable after enough erases, but that’s why there’s redundant data used to store extra bits to implement error correction. But the thing that has the most potential to kill an SSD is corruption of the lookup table to map wear leveling. I don’t know if maybe APFS has any kind of provision that allows for recovery if the mapping is corrupted. Wear leveling is supposed to be transparent to the interface, so I’m not sure if there’s any way to just read out the raw blocks.

The worst thing that can happen to a hard drive is a mechanical failure, but they can be rebuilt. An electrical failure might require a donor board.

Im just wondering when and if something like the Intel/Micron 3D XPoint technology will take off. There’s been talk that it could be used similar to DRAM, but right now Optane drives are too expensive to replace flash. I was thinking maybe as a front end cache for flash SSDs. Those are supposed to be as reliable as DRAM.