In this tutorial I will share the steps to mount NFS share on the client nodes
Overview on NFS
- The Network File System (NFS) is a distributed file system that provides transparent access to remote disks
- Instead of duplicating common directories such as /usr/local on every system, NFS provides a single copy of the directory that is shared by all systems on the network.
- For the user, NFS means that he or she doesn’t have to log into other systems to access files.
- An NFS server is a host that owns one or more filesystems and makes them available on the network;
- NFS clients mount filesystems from one or more servers.
Pre-requisites
Setup NFS exports Server
Configuring NFS Server is not covered as part of this article so I will assume you already a NFS server up and running. In this article we will only cover the NFS client part i.e. to mount NFS share on the client from the server.
In this example I have setup nfs exports on server1 (10.43.138.1) with below configuration
[root@server1 ~]# exportfs -v /ISS <world>(sync,wdelay,hide,no_subtree_check,sec=sys,rw,secure,no_root_squash,no_all_squash)
Install NFS Client
The NFS client package will vary based on the Linux distribution.
On Ubuntu install nfs-common
$ sudo apt install -y nfs-common
On RHEL/CentOS environment install nfs-utils
[root@server2 ~]# yum -y install nfs-utils
Mount NFS File System manually
You can use mount command to mount the NFS file system form remote server to your localhost. The syntax to mount NFS File System on the client node would be:
mount [OPTIONS] NFS_SERVER:/PATH/TO/EXPORTED/DIR /MOUNT_POINT_ON_CLIENT
As per our configuration
NFS_SERVERisserver1(10.43.138.1)/PATH/TO/EXPORTED/DIRis/ISS/MOUNT_POINT_ON_CLIENTis/tmp/logs
So to mount NFS manually we will execute below command on the client i.e. server2 (10.43.138.2)
We need the mount point, so I will create the mount point
[root@server2 ~]# mkdir /tmp/logs
Next mount the NFS file system from server1 on server2
[root@server2 ~]# mount -t nfs 10.43.138.1:/ISS /tmp/logs
Verify if the NFS FS is mounted properly
[root@server2 ~]# df -h /tmp/logs Filesystem Size Used Avail Use% Mounted on 10.43.138.1:/ISS 685G 169G 482G 26% /tmp/logs
Now based on the permission of your NFS share you can access the data of /ISS from server1 on /tmp/logs on server2
Mount NFS File System Persistently
Now with mount command the changes are not persistent and will not survive a reboot. So if you wish to mount your NFS File System after every reboot then you must add this in /etc/fstab or create a systemd unit file to update fstab during reboot.
In this article we will use our traditional fstab to auto-mount the file system. First of all make sure your mount point exists
[root@server2 ~]# mkdir /tmp/logs
Next update /etc/fstab with below content
# NFS_SERVER:/PATH/TO/EXPORTED/DIR /MOUNT_POINT_ON_CLIENT TYPE_OF_FS OPTIONS DUMP PASS
10.10.0.10:/backups /var/backups nfs defaults 0 0
Save and exit the file.
Next to verify if this is working, first un-mount the NFS File System (if in mounted state)
[root@server2 ~]# umount /tmp/logs
Now we will use fstab to mount all the FS available in /etc/fstab. You can safely execute this command and it will not break anything in your environment
[root@server2 ~]# mount -a
Now verify if your NFS File System is properly mounted:
[root@server2 ~]# df -h /tmp/logs Filesystem Size Used Avail Use% Mounted on 10.43.138.1:/ISS 685G 169G 482G 26% /tmp/logs
That's it so next time your client reboots, the NFS File System will be automatically mounted as long as your NFS server is up and running.
Additional Options to mount NFS File System
This section will be applicable based on your NFS server configuration.
For example in the above case if I check for the options using which NFS was mounted
[root@server2 ~]# mount | grep logs 10.43.138.1:/ISS on /tmp/logs type nfs4 (rw,relatime,vers=4.1,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=10.43.138.2,local_lock=none,addr=10.43.138.1)
Here this command tells us that the NFS was mounted using NFSv4 with different rsize, wsize etc. So while performing the mount we have an option to modify these values.
Provide NFS version while mounting the NFS File System
If your NFS server allows you to choose a NFS version for the client mount then you can use -o nfsvers=<ver>, for example to mount using NFSv3
[root@server2 ~]# mount -o nfsvers=3 10.43.138.1:/ISS /tmp/logs
Similarly if your client and server supports you can provide different NFS version.
Setting Block Size to Optimize Transfer Speeds
The mount command options rsize and wsize specify the size of the chunks of data that the client and server pass back and forth to each other. If no rsize and wsize options are specified, the default varies by which version of NFS we are using.
[root@server2 ~]# mount -o rsize=<val>,wsize=<val> 10.43.138.1:/ISS /tmp/logs
Removing NFS Mounts from the Client
To un-mount the NFS mount point you can just use umount command followed by the mount point path
# umount /MOUNT_POINT
OR if you are not aware of the mount point you can also provide the REMOTE_SERVER and REMOTE_DIR PATH i.e.
# umount NFS_SERVER:/PATH/TO/EXPORTED/DIR
So in our example to un-mount our NFS File System we will use
[root@server2 ~]# umount /tmp/logs
OR
[root@server2 ~]# umount 10.43.138.1:/ISS
Next use df or mount command to make sure the NFS FS is not mounted any more.
How to fix "umount: device is busy"
It is possible sometimes you may get this error
umount.nfs4: <mount_point>: device is busy
This is most likely because the mount_point on which your NFS file system is mounted is in use by some process. So either you can find and kill that process or let the process complete.
Kill the process
To kill the process first you should know the process which is occupying the mount_point. You can use
# fuser -m /mount_point
This will give you the PID of the process using the mount_point in below format, in this example my mount_point is /mnt:
[root@server2 ~]# fuser -m /mnt /mnt: 12594
So now you can search for process with PID 12594
[root@server2 ~]# ps -p 12594 PID TTY TIME CMD 12594 pts/0 00:00:00 tail
Now since you know the process which is occupying your NFS mount point, you can either choose to kill it or let the process complete
You can also use lsof to detect the list of process using a file system:
# lsof +f -- <mountpoint or device>
Lazy un-mount
If you choose to ignore the PID and want to let it complete, how would you know when the process is complete?
You will have to constantly monitor the PID to make sure it is complete after which only you can un-mount your NFS File System. SO to overcome this we have something called lazy un-mount where we can trigger umount with -l or --lazy, for example:
[root@server2 ~]# umount /mnt umount.nfs4: /mnt: device is busy
The default umount is not working so let's try lazy un-mount
[root@server2 ~]# umount --lazy /mnt
The exit status is success so our command was executed successfully:
[root@server2 ~]# echo $? 0
So this will detach the file system from the file system hierarchy now, and cleanup all references to the file system as soon as it is not busy anymore. The recommended use-case for umount -l is to prevent hangs on shutdown due to an unreachable network share where a normal umount will hang due to a downed server or a network partition.
[root@server2 ~]# umount --lazy /mnt
Conclusion
In this tutorial we learned about methods to mount NFS shares on client nodes. You can use these steps across different Linux distributions such as Red Hat, Fedora, CentOS, SuSE, Ubuntu, Debian etc. The only difference would be the NFS client package to be installed.
Lastly I hope the steps from the article to mount NFS on Linux was helpful. So, let me know your suggestions and feedback using the comment section.
