Although not mentioned in the description, the slitaz-arm server edition does also include nfs.
I have tested this against a slitaz4 machine and it works.
There is an issue with the init script though (/etc/init.d/nfsd), or actually two.
One: Upon start, the nproc parameter from the file is ignored and
Possible fix:
The script contains a variable OPTION="8" which can set the number of nfs processes to start.
When settings this to another value, you notice it does not work. The value for $OPTION is ignored. Reason for this is the next line in the script:
[ -n "$NFSD_OPTION" ] || OPTION="$NFSD_OPTION"
I was unable to find the meaning of "-n" in the bash shell docu but the xpression seems to evaluate as false. Hence the value of $OPTION gets overwritten by "$NFSD_OPTION" which I do not know how/where to set. Perhaps this was intended to be set in /etc/daemons.conf? Anyway, commenting out this line causes the preset number of processes to be started.
Two: When stopping, processes are not properly killed meaning nfsd keeps running.
When running an nfs server, not all nfsd processes are killed but only the one in the PIDfile.
Even in client mode, the killall lines do not seem to do the job.
Fix:
Three lines were commented:
# killall rpc.mountd
# killall nfsd
# killall lockd
And replaced by:
kill -9 $(pgrep nfsd)
kill $(pgrep rpc.mountd)
lockd exits automatically when the last nfsd process is killed.
I have noticed that the exact same script is also used in slitaz4.
Perhaps it is worth wile to fix this?
/emgi
