In the /etc/rc.d/init.d/netfs script the last step is to do a "mount -a". We've already done a mount all in the rc.sysinit script, why is it here aswell? This reduces the usefullness of this as a seperate script somewhat. Certainly don't exepect my unmounted local file systems to re-appear becuase I've brought up the networked ones. We've gone half the distance by having a seperate netfs script in the first place so why not seperate the functions completly? (personally I think the mounting the non root/boot fs's should be seperate from rc.sysinit to but that's neither here nor there. :))
The reason that the mount -a is there is so that filesystems that might get mounted *under* the network file systems get mounted. (say, /usr is NFS and /usr/local is local.) Admittedly, it's an odd case, but having it there shouldn't hurt anything.