SuSE

Dear all,
in one of our SLES 10 NFS Client there is this statement in its
/etc/fstab:

NFS-Server:/tools /tools nfs hard,bg,intr,noatime 0 0

In a case where a reboot of the NFS Server and this client is necessary
- due to some reasons we have to boot the client first and then it's
NFS Server - we would expect that the client's NFS request is trying
in the background to get the NFS /tools system as long as the NFS server
is up and running.
But that's not the case: We have to do a #mount -a on the client site to get
/tools.
Why?

THX in advance for any help!
Rainer

You may want to look into automount instead of hard (which though explicit, is the default). The hard mount will only matter if the mount happened at all. The "hard" parameter says that if the connection FAILS it will lock until it resumes. It does not state that it will keep trying to mount the filesystem.

If you boot the client before the remote share is available, then you never mounted the remote system. That is why you have to manually mount it later.

With SLES you must activate nfs & nfsboot services to mount nfs @ reboot.