After that we get:
Failed to start load/save random number
Failed to open pack file: read-only file system
…booting continues and mounts disk file systems properly
Failed to write new random seed file: bad file desciptor
overwriting earlier assignment of kernel/sysrq in file …
…later in sequence…
failed to create file/var/log/wtmp: read only file system
…and others of that ilk.
mtab shows:
rootfs / footfs rw 0 0
…and all other temp file systems are rw…
…but /dev/sdf2 is read only. The root fs resides in this partition, and
I can write to it from console.
I had to copy off screen and again onto this form, so there are surely typos.
So far I can get any more diagnostic info and it’s timeconsuming repeatedly to boot,
as this is a server.
I’ve had one trouble after another upgrading from 12.3 to 13.2. Hope someone can help.
Best, jim
You write 13.3 in the title, but there is no 13.3 yet.
According to the rest of your post, I suppose this is 13.2, right?
Well, your / filesystem seems to be read-only, or damaged.
Try to run fsck. As this is the root filesystem, you cannot do that from inside the running system though.
So either boot from some CD/DVD/USB drive, or add “rd.break” to the boot options (press ‘e’ at the boot menu and append this to the line starting with “linux”). This will stop the boot in the initrd before the system is switched to the actual / partition.
You can unmount it then (umount /sysroot) and run fsck.
I’ve had one trouble after another upgrading from 12.3 to 13.2. Hope someone can help.
Well, upgrading from 12.3 to 13.2 is not really supported. Only upgrading to the next version is actually tested.
But it might help if you would describe what problems you have in more detail… 
fsck -l said only that the time was in the future. Running it fixed the boot issue.
I would have thought that the boot process would use temporary file system for
pack file, random #s, etc. I didn’t realize that it required access to a disk
file system.
Thanks much, jim
Well, the initial boot is done from the initrd. But then the actual / is mounted and the system switched to it.
Otherwise you’d need a several GiB big initrd that is an image of the complete / partition which would have to be synchronized with your / partition all the time too…
Normally an fsck is run during boot before / is mounted, but sometimes this cannot fix all problems. And in your case, the problem actually seems to have been a wrong timestamp (“fsck -l said only that the time was in the future”), in which case the automatic check is aborted I think (for safety reasons).