I reinstalled my desktop workstation to replace my Windows/SuSE installation and use autofs to mount home dirs dynamically from my server. But KDE startup took half an hour, dmesg should that statd could not be reached and lockd cannot be reached due to timeout. But rpcinfo showed statd running. It took me two days to figure out that I needed to also install nfs-utils (which installs rpc.statd on the client) and add it to the runlevel. So I suggest that the autofs ebuild includes a useflag "nfs" to depend on nfs-utils and modify the init script to depend on the nfsmount service (probably only if the auto maps include nfs mappings). That would be nice. Reproducible: Always Steps to Reproduce: 1. Compile client and server with kernel nfs support 2. Install nfs-utils on the server to export /home as nfs share 3. Install autofs on the client to import the home share 4. Start KDE or another software that uses file locking on the share Actual Results: Accesses to the server rpc.lockd time out because there is no local rpc.statd running. This makes the KDE startup take half an hour or more with almost no application running correctly. Expected Results: KDE (or other software using file locking) should work as expected: fast startup similar to local home directories.
New maintainer, net-fs/autofs-5.0.3-r1 is now in the tree.
I guess the problem you state is the reason the init-script contains "use nfs". I however do not think it's autofs's responsibility to make sure all necessary mounting mechanisms are installed and loaded (therefore I don't think the nfs USE-flag would be a particularly good idea). Maybe it would be useful to extend the warning associated with the ebuild. It currently only says that portmap needs to be running for nfs mounts, but I guess that is not sufficient?
(In reply to comment #2) > Maybe it would be useful to extend the warning associated with the ebuild. It > currently only says that portmap needs to be running for nfs mounts, but I > guess that is not sufficient? I suppose that would be sufficient.
Actually this issue is very well documented in nfs howto ( http://tldp.org/HOWTO/NFS-HOWTO/client.html ): "To begin using machine as an NFS client, you will need the portmapper running on that machine, and to use NFS file locking, you will also need rpc.statd and rpc.lockd running on both the client and the server." Thus here is nothing autofs specific. But yea, we'll expand elog message to mention this issue with next bump that will happen pretty soon.
elog updated in 5.0.5-r1. Fixed.
*** Bug 380909 has been marked as a duplicate of this bug. ***