[SOLVED] netfs fails to mount samba mountpoint in fstab

Question

Answer:  netfs fails to mount my samba mountpoint because eventhough
network is before netfs, and network is not started in the background, it
still seems to take a while to complete, but it doesn't block netfs from running.
If I put some stalling code in /etc/rc.d/network, and make it wait, then netfs
will mount the shares, because network was allowed to start and netfs can
it can find the servers.
Now, what I want to know is why my network starts up so slowly.
----------------------- Read thread for details ----------------------
This similar setup is working on my old Gentoo setup.
fstab
//tib/public /mnt/network/tib smbfs rw,user,auto,guest 0 0
tib resolves to a server on my LAN.
When I boot my ARCH system, it gets to netfs daemon, and times out trying to
access the server.  I get the message
5711: Connection to tib failed
SMB connection failed
The boot completes, and I log in and get the following
$ ping tib
PING tiburon.mill (192.168.0.36) 56(84) bytes of data.
64 bytes from tiburon.mill (192.168.0.36): icmp_seq=1 ttl=64 time=0.354 ms
64 bytes from tiburon.mill (192.168.0.36): icmp_seq=2 ttl=64 time=0.371 ms
and, I can mount it like this (which reads /etc/fstab)
$ mount //tib/public
$ ls /mnt/network/tib/ | wc -l
11
It seems like ARCH isn't able to use the network until sometime after this point in the boot.  Here's my rc.conf daemons
DAEMONS=(syslog-ng network netfs crond sshd alsa)
I tried putting netfs last, but that made no difference.  I also tried to load netfs in the background without benefit.
Note:  cifs doesn't work either; and I get more descriptive errors
:: Mounting Network Filesystems
mount error: count not find target server. TCP name tib/public not found
No ip address specified and hostname not found.
OK, that clear enough.  But, the hostname is found right when the boot process finishes and I get a prompt.
And, I've tried the ip address, and FQDN too.
Any ideas?
Last edited by ristretto (2007-07-24 05:04:57)

Answer 1

ok, gave it a try.
//tib/public /mnt/network/tib cifs rw,user,noauto,guest 0 0
$ cat /etc/rc.d/local
#!/bin/bash
mount //tib/public
$ grep DAE /etc/rc.conf
# DAEMONS
DAEMONS=(syslog-ng network netfs crond sshd alsa local)
Note local at the end.
Now, all the daemons load up quick, and then local goes and times out, and I
get the same error.
mount error: count not find target server. TCP name tib/public not found
No ip address specified and hostname not found.
putting mount in .bash_profile is a bit of a security issue (though not really on my setup)
but, more importantly, it mounts each time I log in to that account (say each ssh window to the box.)
I'm thinking it's a bug in the initialization process somewhere?  I've never had this problem on other distros.
What thinks you?