Red Hat Bugzilla – Bug 189101
rpc.lockd not locked down correctly by /etc/sysconfig/nfs
Last modified: 2007-11-30 17:11:30 EST
Description of problem:
rpc.lockd is not locked down correctly by the configuration file
/etc/sysconfig/nfs. For example:
Won't lock down rpc.lockd after a boot. However it typically will if you issue
a "service nfs restart" command.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Create above /etc/sysconfig/nfs file.
2. Reboot the system.
3. Issue "rpcinfo -p" command.
4. Issue "service nfs restart" command.
5. Issue "rpcinfo -p" command again and observe the difference.
rpc.lockd not locked down after a boot.
It should be locked down.
The problem is that although the control script /etc/rc.d/init.d/nfs references
the variables LOCKD_TCPPORT and LOCKD_UDPPORT, the /etc/rc.d/init.d/nfslock
script does not. It seems what is in the nfs script isn't good enough....
perhaps because at the time the nfs script is trying to get rpc.lockd to lock
down its ports, rpc.lockd isn't running yet (or something like that as it works
with a restart).
This problem is in RHEL4 as well but was fixed with a patch. The same patch
fixes FC5 also. See bug 162133 for details.
fixed in nfs-utils-1.0.8-3