Jump to contentJump to page navigation: previous page [access key p]/next page [access key n]
Reference / Services / Sharing file systems with NFS
Applies to openSUSE Leap 15.4

22 Sharing file systems with NFS

  

The Network File System (NFS) is a protocol that allows access to files on a server in a manner similar to accessing local files.

openSUSE Leap installs NFS v4.2, which introduces support for sparse files, file pre-allocation, server-side clone and copy, application data block (ADB), and labeled NFS for mandatory access control (MAC) (requires MAC on both client and server).

22.1 Overview

  

The Network File System (NFS) is a standardized, well-proven and widely supported network protocol that allows sharing files between separate hosts.

The Network Information Service (NIS) can be used to have centralized user management in the network. Combining NFS and NIS allows using file and directory permissions for access control in the network. NFS with NIS makes a network transparent to the user.

In the default configuration, NFS completely trusts the network and thus any machine that is connected to a trusted network. Any user with administrator privileges on any computer with physical access to any network the NFS server trusts can access any files that the server makes available.

Often, this level of security is perfectly satisfactory, such as when the network that is trusted is truly private, often localized to a single cabinet or machine room, and no unauthorized access is possible. In other cases, the need to trust a whole subnet as a unit is restrictive, and there is a need for more fine-grained trust. To meet the need in these cases, NFS supports various security levels using the Kerberos infrastructure. Kerberos requires NFSv4, which is used by default. For details, see Book “Security and Hardening Guide”, Chapter 6 “Network authentication with Kerberos”.

The following are terms used in the YaST module.

Exports

A directory exported by an NFS server, which clients can integrate into their systems.

NFS client

The NFS client is a system that uses NFS services from an NFS server over the Network File System protocol. The TCP/IP protocol is already integrated into the Linux kernel; there is no need to install any additional software.

NFS server

The NFS server provides NFS services to clients. A running server depends on the following daemons: nfsd (worker), idmapd (ID-to-name mapping for NFSv4, needed for certain scenarios only), statd (file locking), and mountd (mount requests).

NFSv3

NFSv3 is the version 3 implementation, the “old” stateless NFS that supports client authentication.

NFSv4

NFSv4 is the new version 4 implementation that supports secure user authentication via Kerberos. NFSv4 requires one single port only and thus is better suited for environments behind a firewall than NFSv3.

The protocol is specified as https://datatracker.ietf.org/doc/html/rfc3530.

pNFS

Parallel NFS, a protocol extension of NFSv4. Any pNFS clients can directly access the data on an NFS server.

Important
Important: Need for DNS

In principle, all exports can be made using IP addresses only. To avoid timeouts, you need a working DNS system. DNS is necessary at least for logging purposes, because the mountd daemon does reverse lookups.

22.2 Installing NFS server

  

The NFS server is not part of the default installation. To install the NFS server using YaST, choose Software › Software Management, select Patterns, and enable the File Server option in the Server Functions section. Click Accept to install the required packages.

Like NIS, NFS is a client/server system. However, a machine can be both—it can supply file systems over the network (export) and mount file systems from other hosts (import).

Note
Note: Mounting NFS volumes locally on the exporting server

Mounting NFS volumes locally on the exporting server is not supported on openSUSE Leap.

22.3 Configuring NFS server

  

Configuring an NFS server can be done either through YaST or manually. For authentication, NFS can also be combined with Kerberos.

22.3.1 Exporting file systems with YaST

  

With YaST, turn a host in your network into an NFS server—a server that exports directories and files to all hosts granted access to it or to all members of a group. Thus, the server can also provide applications without installing the applications locally on every host.

To set up such a server, proceed as follows:

Procedure 22.1: Setting up an NFS server
  

  1. Start YaST and select Network Services › NFS Server; see Figure 22.1, “NFS server configuration tool”. You may be prompted to install additional software.

    NFS server configuration tool
    Figure 22.1: NFS server configuration tool
      

  2. Click the Start radio button.

  3. If firewalld is active on your system, configure it separately for NFS (see Book “Security and Hardening Guide”, Chapter 23 “Masquerading and firewalls”, Section 23.4 “firewalld”). YaST does not yet have complete support for firewalld, so ignore the "Firewall not configurable" message and continue.

    When configuring firewalld rules, add nfs or nfs service with the port value of 2049 for both TCP and UDP. Also add the mountd service with the port value of 20048 for both TCP and UDP.

  4. Check whether you want to Enable NFSv4. If you deactivate NFSv4, YaST will only support NFSv3. For information about enabling NFSv2, see Note: NFSv2.

    1. If NFSv4 is selected, additionally enter the appropriate NFSv4 domain name. This parameter is used by the idmapd daemon that is required for Kerberos setups or if clients cannot work with numeric user names. Leave it as localdomain (the default) if you do not run idmapd or do not have any special requirements. For more information on the idmapd daemon, see /etc/idmapd.conf.

  5. Click Enable GSS Security if you need secure access to the server. A prerequisite for this is to have Kerberos installed on your domain and to have both the server and the clients kerberized. Click Next to proceed with the next configuration dialog.

  6. Click Add Directory in the upper half of the dialog to export your directory.

  7. If you have not configured the allowed hosts already, another dialog for entering the client information and options pops up automatically. Enter the host wild card (usually you can leave the default settings as they are).

    There are four possible types of host wild cards that can be set for each host: a single host (name or IP address), netgroups, wild cards (such as * indicating all machines can access the server), and IP networks.

    For more information about these options, see the exports man page.

  8. Click Finish to complete the configuration.

22.3.2 Exporting file systems manually

  

The configuration files for the NFS export service are /etc/exports and /etc/sysconfig/nfs. In addition to these files, /etc/idmapd.conf is needed for the NFSv4 server configuration with kerberized NFS or if the clients cannot work with numeric user names.

To start or restart the services, run the command systemctl restart nfsserver. This also restarts the RPC port mapper that is required by the NFS server.

To make sure the NFS server always starts at boot time, run sudo systemctl enable nfsserver.

Note
Note: NFSv4

NFSv4 is the latest version of the NFS protocol available on openSUSE Leap. Configuring directories for export with NFSv4 is now the same as with NFSv3.

On openSUSE prior to Leap, the bind mount in /etc/exports was mandatory. It is still supported, but now deprecated.

/etc/exports

The /etc/exports file contains a list of entries. Each entry indicates a directory that is shared and how it is shared. A typical entry in /etc/exports consists of: 

/SHARED/DIRECTORY   HOST(OPTION_LIST)

For example: 

/export/data   192.168.1.2(rw,sync)

Here the IP address 192.168.1.2 is used to identify the allowed client. You can also use the name of the host, a wild card indicating a set of hosts (*.abc.com, *, etc.), or netgroups (@my-hosts).

For a detailed explanation of all options and their meanings, refer to the man page of /etc/exports: (man exports).

In case you have modified /etc/exports while the NFS server was running, you need to restart it for the changes to become active: sudo systemctl restart nfsserver.

/etc/sysconfig/nfs

The /etc/sysconfig/nfs file contains a few parameters that determine NFSv4 server daemon behavior. It is important to set the parameter NFS4_SUPPORT to yes (default). NFS4_SUPPORT determines whether the NFS server supports NFSv4 exports and clients.

In case you have modified /etc/sysconfig/nfs while the NFS server was running, you need to restart it for the changes to become active: sudo systemctl restart nfsserver.

Tip
Tip: Mount options

On openSUSE prior to Leap, the --bind mount in /etc/exports was mandatory. It is still supported, but now deprecated. Configuring directories for export with NFSv4 is now the same as with NFSv3.

Note
Note: NFSv2

If NFS clients still depend on NFSv2, enable it on the server in /etc/sysconfig/nfs by setting: 

NFSD_OPTIONS="-V2"
MOUNTD_OPTIONS="-V2"

After restarting the service, check whether version 2 is available with the command: 

> cat /proc/fs/nfsd/versions
+2 +3 +4 +4.1 +4.2
/etc/idmapd.conf

The idmapd daemon is only required if Kerberos authentication is used or if clients cannot work with numeric user names. Linux clients can work with numeric user names since Linux kernel 2.6.39. The idmapd daemon does the name-to-ID mapping for NFSv4 requests to the server and replies to the client.

If required, idmapd needs to run on the NFSv4 server. Name-to-ID mapping on the client will be done by nfsidmap provided by the package nfs-client.

Make sure that there is a uniform way in which user names and IDs (UIDs) are assigned to users across machines that might be sharing file systems using NFS. This can be achieved by using NIS, LDAP, or any uniform domain authentication mechanism in your domain.

The parameter Domain must be set the same for both client and server in the /etc/idmapd.conf file. If you are not sure, leave the domain as localdomain in the server and client files. A sample configuration file looks like the following: 

[General]
Verbosity = 0
Pipefs-Directory = /var/lib/nfs/rpc_pipefs
Domain = localdomain

[Mapping]
Nobody-User = nobody
Nobody-Group = nobody

To start the idmapd daemon, run systemctl start nfs-idmapd. In case you have modified /etc/idmapd.conf while the daemon was running, you need to restart it for the changes to become active: systemctl start nfs-idmapd.

For more information, see the man pages of idmapd and idmapd.conf (man idmapd and man idmapd.conf).

22.3.3 NFS with Kerberos

  

To use Kerberos authentication for NFS, Generic Security Services (GSS) must be enabled. Select Enable GSS Security in the initial YaST NFS Server dialog. You must have a working Kerberos server to use this feature. YaST does not set up the server but only uses the provided functionality. To use Kerberos authentication in addition to the YaST configuration, complete at least the following steps before running the NFS configuration:

  1. Make sure that both the server and the client are in the same Kerberos domain. They must access the same KDC (Key Distribution Center) server and share their krb5.keytab file (the default location on any machine is /etc/krb5.keytab). For more information about Kerberos, see Book “Security and Hardening Guide”, Chapter 6 “Network authentication with Kerberos”.

  2. Start the gssd service on the client with systemctl start rpc-gssd.service.

  3. Start the svcgssd service on the server with systemctl start rpc-svcgssd.service.

Kerberos authentication also requires the idmapd daemon to run on the server. For more information, refer to /etc/idmapd.conf.

For more information about configuring kerberized NFS, refer to the links in Section 22.6, “More information”.

22.4 Configuring clients

  

To configure your host as an NFS client, you do not need to install additional software. All needed packages are installed by default.

22.4.1 Importing file systems with YaST

  

Authorized users can mount NFS directories from an NFS server into the local file tree using the YaST NFS client module. Proceed as follows:

Procedure 22.2: Importing NFS directories
  

  1. Start the YaST NFS client module.

  2. Click Add in the NFS Shares tab. Enter the host name of the NFS server, the directory to import, and the mount point at which to mount this directory locally.

  3. When using NFSv4, select Enable NFSv4 in the NFS Settings tab. Additionally, the NFSv4 Domain Name must contain the same value as used by the NFSv4 server. The default domain is localdomain.

  4. To use Kerberos authentication for NFS, GSS security must be enabled. Select Enable GSS Security.

  5. Enable Open Port in Firewall in the NFS Settings tab if you use a firewall and want to allow access to the service from remote computers. The firewall status is displayed next to the check box.

  6. Click OK to save your changes.

The configuration is written to /etc/fstab and the specified file systems are mounted. When you start the YaST configuration client at a later time, it also reads the existing configuration from this file.

Tip
Tip: NFS as a root file system

On (diskless) systems where the root partition is mounted via network as an NFS share, you need to be careful when configuring the network device with which the NFS share is accessible.

When shutting down or rebooting the system, the default processing order is to turn off network connections then unmount the root partition. With NFS root, this order causes problems as the root partition cannot be cleanly unmounted as the network connection to the NFS share is already deactivated. To prevent the system from deactivating the relevant network device, open the network device configuration tab as described in Section 13.4.1.2.5, “Activating the network device” and choose On NFSroot in the Device Activation pane.

22.4.2 Importing file systems manually

  

The prerequisite for importing file systems manually from an NFS server is a running RPC port mapper. The nfs service takes care to start it properly; thus, start it by entering systemctl start nfs as root. Then remote file systems can be mounted in the file system just like local partitions, using the mount: 

> sudo mount HOST:REMOTE-PATHLOCAL-PATH

To import user directories from the nfs.example.com machine, for example, use: 

> sudo mount nfs.example.com:/home /home

To define a count of TCP connections that the clients make to the NFS server, you can use the nconnect option of the mount command. You can specify any number between 1 and 16, where 1 is the default value if the mount option has not been specified.

The nconnect setting is applied only during the first mount process to the particular NFS server. If the same client executes the mount command to the same NFS server, all already established connections will be shared—no new connection will be established. To change the nconnect setting, you have to unmount all client connections to the particular NFS server. Then you can define a new value for the nconnect option.

You can find the value of nconnect that is in currently in effect in the output of the mount, or in the file /proc/mounts. If there is no value for the mount option, then the option has not been used during mounting and the default value of 1 is in use.

Note
Note: Different number of connections than defined by nconnect

As you can close and open connections after the first mount, the actual count of connections does not necessarily have to be the same as the value of nconnect.

22.4.2.1 Using the automount service

  

The autofs daemon can be used to mount remote file systems automatically. Add the following entry to the /etc/auto.master file: 

/nfsmounts /etc/auto.nfs

Now the /nfsmounts directory acts as the root for all the NFS mounts on the client if the auto.nfs file is filled appropriately. The name auto.nfs is chosen for the sake of convenience—you can choose any name. In auto.nfs add entries for all the NFS mounts as follows: 

localdata -fstype=nfs server1:/data
nfs4mount -fstype=nfs4 server2:/

Activate the settings with systemctl start autofs as root. In this example, /nfsmounts/localdata, the /data directory of server1, is mounted with NFS and /nfsmounts/nfs4mount from server2 is mounted with NFSv4.

If the /etc/auto.master file is edited while the service autofs is running, the automounter must be restarted for the changes to take effect with systemctl restart autofs.

22.4.2.2 Manually editing /etc/fstab

  

A typical NFSv3 mount entry in /etc/fstab looks like this: 

nfs.example.com:/data /local/path nfs rw,noauto 0 0

For NFSv4 mounts, use nfs4 instead of nfs in the third column: 

nfs.example.com:/data /local/pathv4 nfs4 rw,noauto 0 0

The noauto option prevents the file system from being mounted automatically at start-up. If you want to mount the respective file system manually, it is possible to shorten the mount command specifying the mount point only: 

> sudo mount /local/path
Note
Note: Mounting at start-up

If you do not enter the noauto option, the init scripts of the system will handle the mount of those file systems at start-up.

22.4.3 Parallel NFS (pNFS)

  

NFS is one of the oldest protocols, developed in the 1980s. As such, NFS is usually sufficient if you want to share small files. However, when you want to transfer big files or many clients want to access data, an NFS server becomes a bottleneck and has a significant impact on the system performance. This is because files are quickly getting bigger, whereas the relative speed of Ethernet has not fully kept pace.

When you request a file from a regular NFS server, the server looks up the file metadata, collects all the data, and transfers it over the network to your client. However, the performance bottleneck becomes apparent no matter how small or big the files are:

  • With small files, most of the time is spent collecting the metadata.

  • With big files, most of the time is spent on transferring the data from server to client.

pNFS, or parallel NFS, overcomes this limitation as it separates the file system metadata from the location of the data. As such, pNFS requires two types of servers:

  • A metadata or control server that handles all the non-data traffic

  • One or more storage server(s) that hold(s) the data

The metadata and the storage servers form a single, logical NFS server. When a client wants to read or write, the metadata server tells the NFSv4 client which storage server to use to access the file chunks. The client can access the data directly on the server.

openSUSE Leap supports pNFS on the client side only.

22.4.3.1 Configuring pNFS client with YaST

  

Proceed as described in Procedure 22.2, “Importing NFS directories”, but click the pNFS (v4.2) check box and optionally NFSv4 share. YaST will do all the necessary steps and will write all the required options in the file /etc/exports.

22.4.3.2 Configuring pNFS client manually

  

Refer to Section 22.4.2, “Importing file systems manually” to start. Most of the configuration is done by the NFSv4 server. For pNFS, the only difference is to add the minorversion option and the metadata server MDS_SERVER to your mount command: 

> sudo mount -t nfs4 -o minorversion=1 MDS_SERVER MOUNTPOINT

To help with debugging, change the value in the /proc file system: 

> sudo echo 32767 > /proc/sys/sunrpc/nfsd_debug
> sudo echo 32767 > /proc/sys/sunrpc/nfs_debug

22.5 Managing Access Control Lists over NFSv4

  

There is no single standard for Access Control Lists (ACLs) in Linux beyond the simple read, write, and execute (rwx) flags for user, group, and others (ugo). One option for finer control is the Draft POSIX ACLs, which were never formally standardized by POSIX. Another is the NFSv4 ACLs, which were designed to be part of the NFSv4 network file system with the goal of making something that provided reasonable compatibility between POSIX systems on Linux and WIN32 systems on Microsoft Windows.

NFSv4 ACLs are not sufficient to correctly implement Draft POSIX ACLs so no attempt has been made to map ACL accesses on an NFSv4 client (such as using setfacl).

When using NFSv4, Draft POSIX ACLs cannot be used even in emulation and NFSv4 ACLs need to be used directly; that means while setfacl can work on NFSv3, it cannot work on NFSv4. To allow NFSv4 ACLs to be used on an NFSv4 file system, SUSE Linux Enterprise Server provides the nfs4-acl-tools package, which contains the following:

  • nfs4-getfacl

  • nfs4-setfacl

  • nfs4-editacl

These operate in a generally similar way to getfacl and setfacl for examining and modifying NFSv4 ACLs. These commands are effective only if the file system on the NFS server provides full support for NFSv4 ACLs. Any limitation imposed by the server will affect programs running on the client in that some particular combinations of Access Control Entries (ACEs) might not be possible.

It is not supported to mount NFS volumes locally on the exporting NFS server.

Additional Information

For information, see Introduction to NFSv4 ACLs at http://wiki.linux-nfs.org/wiki/index.php/ACLs#Introduction_to_NFSv4_ACLs.

22.6 More information

  

In addition to the man pages of exports, nfs, and mount, information about configuring an NFS server and client is available in /usr/share/doc/packages/nfsidmap/README. For further documentation online, refer to the following Web sites:

  • For general information about network security, refer to Book “Security and Hardening Guide”, Chapter 23 “Masquerading and firewalls”.

  • Refer to Section 23.4, “Auto-mounting an NFS share” if you need to automatically mount NFS exports.

  • For more details about configuring NFS by using AutoYaST, refer to Book “AutoYaST Guide”, Chapter 4 “Configuration and installation options”, Section 4.19 “NFS client and server”.

  • For instructions about securing NFS exports with Kerberos, refer to Book “Security and Hardening Guide”, Chapter 6 “Network authentication with Kerberos”, Section 6.6 “Kerberos and NFS”.

  • Find the detailed technical documentation online at SourceForge.

22.7 Gathering information for NFS troubleshooting

  

22.7.1 Common troubleshooting

  

In some cases, you can understand the problem in your NFS by reading the error messages produced and looking into the /var/log/messages file. However, in many cases, the information provided by the error messages and in /var/log/messages is not detailed enough. In these cases, most NFS problems can be best understood through capturing network packets while reproducing the problem.

Clearly define the problem. Examine the problem by testing the system in a variety of ways and determining when the problem occurs. Isolate the simplest steps that lead to the problem. Then try to reproduce the problem as described in the procedure below.

Procedure 22.3: Reproducing the problem
  

  1. Capture network packets. On Linux, you can use the tcpdump command, which is supplied by the tcpdump package.

    An example of tcpdump syntax follows: 

    tcpdump -s0 -i eth0 -w /tmp/nfs-demo.cap host x.x.x.x

    Where:

    s0

    Prevents packet truncation

    eth0

    Should be replaced with the name of the local interface which the packets will pass through. You can use the any value to capture all interfaces at the same time, but usage of this attribute often results in inferior data as well as confusion in analysis.

    w

    Designates the name of the capture file to write.

    x.x.x.x

    Should be replaced with the IP address of the other end of the NFS connection. For example, when taking a tcpdump at the NFS client side, specify the IP address of the NFS Server, and vice versa.

    Note
    Note

    In some cases, capturing the data at either the NFS client or NFS server is sufficient. However, in cases where end-to-end network integrity is in doubt, it is often necessary to capture data at both ends.

    Do not shut down the tcpdump process and proceed to the next step.

  2. (Optional) If the problem occurs during execution of the nfs mount command itself, you can try to use the high-verbosity option (-vvv) of the nfs mount command to get more output.

  3. (Optional) Get an strace of the reproduction method. An strace of reproduction steps records exactly what system calls were made at exactly what time. This information can be used to further determine on which events in the tcpdump you should focus.

    For example, if you found out that executing the command mycommand --param was failing on an NFS mount, then you could strace the command with: 

    strace -ttf -s128 -o/tmp/nfs-strace.out mycommand --param

    In case you do not get any strace of the reproduction step, note the time when the problem was reproduced. Check the /var/log/messages log file to isolate the problem.

  4. Once the problem has been reproduced, stop tcpdump running in your terminal by pressing CTRL–c. If the strace command resulted in a hang, also terminate the strace command.

  5. An administrator with experience in analyzing packet traces and strace data can now inspect data in /tmp/nfs-demo.cap and /tmp/nfs-strace.out.

22.7.2 Advanced NFS debugging

  
Important
Important: Advanced debugging is for experts

Please bear in mind that the following section is intended only for skilled NFS administrators who understand the NFS code. Therefore, perform the first steps described in Section 22.7.1, “Common troubleshooting” to help narrow down the problem and to inform an expert about which areas of debug code (if any) might be needed to learn deeper details.

There are various areas of debug code that can be enabled to gather additional NFS-related information. However, the debug messages are quite cryptic and the volume of them can be so large that the use of debug code can affect system performance. It may even impact the system enough to prevent the problem from occurring. In the majority of cases, the debug code output is not needed, nor is it typically useful to anyone who is not highly familiar with the NFS code.

22.7.2.1 Activating debugging with rpcdebug

  

The rpcdebug tool allows you to set and clear NFS client and server debug flags. In case the rpcdebug tool is not accessible in your SLE, you can install it from the package nfs-client or nfs-kernel-server for the NFS server.

To set debug flags, run: 

rpcdebug -m module -s flags

To clear the debug flags, run: 

rpcdebug -m module -c flags

where module can be:

nfsd

Debug for the NFS server code

nfs

Debug for the NFS client code

nlm

Debug for the NFS Lock Manager, at either the NFS client or NFS server. This only applies to NFS v2/v3.

rpc

Debug for the Remote Procedure Call module, at either the NFS client or NFS server.

For information on detailed usage of the rpcdebug command, refer to the manual page: 

man 8 rpcdebug

22.7.2.2 Activating debug for other code upon which NFS depends

  

NFS activities may depend on other related services, such as the NFS mount daemon—rpc.mountd. You can set options for related services within /etc/sysconfig/nfs.

For example, /etc/sysconfig/nfs contains the parameter: 

MOUNTD_OPTIONS=""

To enable the debug mode, you have to use the -d option followed by any of the values: all, auth, call, general, or parse.

For example, the following code enables all forms of rpc.mountd logging: 

MOUNTD_OPTIONS="-d all"

For all available options refer to the manual pages: 

man 8 rpc.mountd

After changing /etc/sysconfig/nfs, services need to be restarted: 

systemctl restart nfsserver  # for nfs server related changes
systemctl restart nfs  # for nfs client related changes