Nmclient object error after running puppet

It turns out that the problem wasn't the main RAID1 partition, it was the other partitions.

In the first place, I shouldn't have used swap partitions. That was just dumb. Even if that had worked, it still probably would have made the system vulnerable to a failure if a disk drive developed a bad block within the swap partition. It's obviously better to use a swap file on the RAID1 partition; I don't know what I was thinking.

However, the "extra" ext3 md1 partition was also a problem. I don't know why.

Once I removed the references to the other partitions (the two swap partitions and the ext3 md1 partition) from /etc/fstab, the system would boot up just fine with one drive, running the RAID1 array in degraded mode, like I wanted it to.

After I shut down and reinstalled the missing drive, I started the machine again, and it was still running with just one drive. But I did "mdadm --add" to add the missing drive back, and its state went to "spare rebuilding" for a while, and then to "active."

In other words, it's working perfectly.

I finally figured this out. You need to give the docker container the capability to use ptrace. Simply adding --cap-add SYS_PTRACE to the docker run command fixed this problem for me.