I have a machine that will not boot up off of NVMe so I used a small SATA device to jumpstart the box.
When I installed FreeBSD on it I put /usr on the NVMe.
My plans were to build NanoBSD on this box and I wanted to take advantage of the speedy disk.
So it seemed logical to put all of /usr on it.
In fact all I really needed on it were /usr/src, usr/ports and usr/obj for NanoBSD.
With all of /usr mount on it I found problems when I disconnected the device, because so much more is in /usr
So what should have been my strategy here?
3 separate partitions on the NVMe drive and each one mounted at their respective /usr/ locations?
(With FAILOK set in fstab)
I am retasking the box and trying to take lessons from it.
I realize the traditional UNIX way was separate partitions for many of these directories. So I am lacking in the fundamentals.
Separate partitions for sub-directories is foreign to me.
When I installed FreeBSD on it I put /usr on the NVMe.
My plans were to build NanoBSD on this box and I wanted to take advantage of the speedy disk.
So it seemed logical to put all of /usr on it.
In fact all I really needed on it were /usr/src, usr/ports and usr/obj for NanoBSD.
With all of /usr mount on it I found problems when I disconnected the device, because so much more is in /usr
So what should have been my strategy here?
3 separate partitions on the NVMe drive and each one mounted at their respective /usr/ locations?
(With FAILOK set in fstab)
I am retasking the box and trying to take lessons from it.
I realize the traditional UNIX way was separate partitions for many of these directories. So I am lacking in the fundamentals.
Separate partitions for sub-directories is foreign to me.