nfs client configuration : system trys to mount the mounts more than once
So I was trying not to install 10 copies of centos 6.2 on some new machines - ended up cloning one installed copy to use and putting that copy on all of the systems - taking care to change IP addresses and hostnames... but for some reason i see the following message for several of my mounts on nearly all of the systems:
mount.nfs : <some local nfs path used for mount> is busy or already mounted
note this is not an "error" as my mounts seem to be working, just trying to keep things nice and clean so i'd like to figure out why the system is appearing to try to mount the mounts more than once.
Ive looked at:
/etc/fstab and that looks ok
mtab and the stuff in proc are dynamic tables (so nothing to do there)
what else can be causing these messages to show up on boot?
whats somewhat interesting is that when i go to the directory of the mount
(from fstab on ironboard)
bobbin:/shared /shared nfs rsize=8192,wsize=8192,timeo=14,intr
so on my local box ironboard in /shared i see all the stuff mounted from bobbin
However... if in terminal i do "mount" i do not see the mount...
I have used NFS in debian. The following package should be installed on server
$ sudo apt-get install nfs-kernel-server
and following package should be installed on client
$ $ sudo apt-get install nfs-common
The directory what you want to share on server and access on client, should be mentioned in /etc/exports.
This file will not be on client's system, it will be on server because you have to give permission on server to access files.
Now mount that file on client which is shared on server.
$ sudo mount 192.168.0.101:/root/Desktop/process /home/user/test
The process directory of server is mounted on user's account at given path.
For RHEL, user yum instead of apt-get to install server and client packages.
Um ok perhaps I wasn't real clear, so we have a server that was working and that some of the clients still connect to without issue. So dont believe this is a server issue as its the same server we've had for some time.
Also as i said in the original post - the mounts do work on the new machines that upgraded (the client does connect). But sometimes i see strange NFS issues and the "busy or already mounted" messages. And being that the systems are clones i suspect something is still amiss in the networking department. Perhaps stale nfs locks or something else copied in the cloning process that might upset NFS
I suspect having several machines with the exact same UUID connected to an NFS server is the problem but not sure... but my SGE wont launch jobs on it for some reason and i dont know why.
|All times are GMT -5. The time now is 12:38 PM.|