Anyone else having problems with rpc.lockd not starting at boot
time? There's a check in the /etc/rc.d/init.d/nfslock file, in the form
# don't start lockd if kernel-linus.
if ! grep -q mdk /proc/version;then
echo -n "Starting NFS lockd: "
daemon rpc.lockd
echo
fi
The comment states that this should NOT run lockd if you're running
kernel-linus. Forgive me if I'm wrong here, but doesn't the "if"
statement do just the opposite; no lockd if kernel-x.x.xx-xmdk,
otherwise
run lockd if kernel-linus? Therefore, no working nfs locking if you're
running an mdk kernel. Is there some reason I don't know about for
ditching NFS locking? I've fixed the nfslock script here and gotten
lockd
running since we really can't do without it (too many *NIX boxen of
sundry
falvors), but I was just wondering if this was a typo or if I'm actually
missing something crucial. Also, I should mention that, earlier in the
nfslock script, these two tests appear;
[ -x /sbin/rpc.lockd ] || exit 0
[ -x /sbin/rpc.statd ] || exit 0
Note that, at least on the 7.1 installs I've done so far, all the rpc.*
daemons are in /usr/sbin, not /sbin, another test that bombs out the
script. Small problems, admittedly, but not so trivial when the number
of
machines gets large enough.
<2:00AM coffee-deficient complaint of the day below>
Otherwise, I like 7.1. The only complaint I have is the new menu
structure; rather caught me by suprise. It works fine, no complaints
there, but since we have a lot of off-the-wall software, I'd spent a lot
of time generating custom KDE and Gnome menus that were consistent
across
the network. Needless to say, it didn't really survive the upgrade in a
seamless fashion, but it's fixed now. Some warning that I was about to
trash 14 months work everytime I added a package would have been nice,
but
it's my fault anyway...
--
Lee Burnside -- [EMAIL PROTECTED]
http://www.phys.ttu.edu/~tljlb/
"Many a good hanging prevents a bad marriage."
--The Bard
