is there already a good wiki on how to start with NFS on Devuan?

I do not see any big difference between Debian and Devuan regarding the implementation of a nfs server and client. The guides you can find on the internet do apply for both in general.

I have started a file server on Sarge, first of all with Samba for the Windows PCs on my family network. When I started with dual booting with Linux on my home PC, I discovered that I could export the shares I had prepared for Samba easily for nfs too. I started off with nfs(3), and moved to nfs4 some years later. There are a few differences in the way you export the shares, but there is no relation to any init system.

On the client side, I wasn't successful with mounting the shares via fstab due to the fact that the file server was put to sleep at night, and it would wake up past the client, and it needed some additional time for decrypting the shares. So I created a script that is invoked by cron @reboot. The script checks if the server is online, then if the shares are exported, and then it mounts the shares. When the server is off the operation of my workstation isn't affected, just the shares aren't there. I am still using this proven and reliable approach nowadays.

Add this at the bottom of /etc/auto.master:

/mnt	/etc/auto.nfs

Create /etc/auto.nfs containing something like:

<share-name>    /path/to/<your-symlink-name>

Make the symlink and restart autofs:

ln -s /mnt/<share-name> /path/to/<your-symlink-name>
service autofs restart

- your-symlink-name can be the same or different from share-name.
- If you're already using /mnt for something else, you can mount the share somewhere else. Just make an empty directory to be used as the mountpoint.
- Don't use fstab for any part of this.

the "AI Overview" told me how to automount NFS on Devuan, but it came with systemctl commands through the road, so led my bit astray.

RPCMOUNTDOPTS="--manage-gids --debug-all"

Removing the dash in second option changing it to "--debug all" made the script work.

rolfie

# service nfs-kernel-server restart
Stopping NFS kernel daemon: mountd nfsd.
Unexporting directories for NFS kernel daemon....
Exporting directories for NFS kernel daemon....
Starting NFS kernel daemon: nfsd mountdUsage: rpc.mountd [-F|--foreground] [-h|--help] [-v|--version] [-d kind|--debug kind]
	[-o num|--descriptors num]
	[-p|--port port] [-V version|--nfs-version version]
	[-N version|--no-nfs-version version] [-n|--no-tcp]
	[-H prog |--ha-callout prog] [-r |--reverse-lookup]
	[-s|--state-directory-path path] [-g|--manage-gids]
	[-t num|--num-threads=num] [-u|--no-udp]
 failed!
# showmount -e fs090
clnt_create: RPC: Program not registered 

I don't know how to deal with this situation.

Background info: Xeon 4 core CPU on X99-E WS, 32G RAM, 256G Plextor nvme as system disk, LSI9271 supplying a RAID6 with about 5TB space, two more disks for single shares. Installed from Chimaera netinstall, encrypted LVM, openrc as init, manual installation of xorg, lightdm, Mate desktop, nfs-common, nfs-kernel-server. Latest stock kernel and backports kernels available, 5.15 is the default. It does not matter which one I do boot. The RAID6 is nicely decrypted and mounted, also the two other disks. The export file is a copy of the original one from a backup. Most of the setup I have taken from my notes. Everything should work.

The above error is shown as well during boot as when I try to restart the nfs-server. Also showmount throws an error. The duck does not give hints to this error.

Also tried to drill down how the nfs-server is started. The script in /etc/rc0.d does not contain this specific error printout. How can I find the root cause for the startup failure?

Thanks, rolfie