Difference between revisions of "NFS server inside container"

From OpenVZ Virtuozzo Containers Wiki
Jump to: navigation, search
(link to NFS)
(Improve the grammar a little)
Line 17: Line 17:
 
== User-space NFS server ==
 
== User-space NFS server ==
 
Advantage of user-space NFS server is that it can be ran without any support in the kernel.
 
Advantage of user-space NFS server is that it can be ran without any support in the kernel.
Also if it crashes — there is no crash of a system: just one process crashes, not kernel!
+
Also if it crashes — there is no crash of the system: just one process crashes, not the kernel!
The disadvantage of user-space NFS server is its productivity: no one can be faster then in-kernel implementation.
+
The disadvantage of user-space NFS server is its productivity: no one can be faster than in-kernel implementation.
  
 
One well-known implementation of NFS server is "The LINUX User-Space NFS Server" by Olaf Kirch.
 
One well-known implementation of NFS server is "The LINUX User-Space NFS Server" by Olaf Kirch.
 
Some Linux distributions contain this package: Debian Sarge (<code>nfs-user-server</code>), OpenSUSE 10.0 (<code>nfs-server</code>).
 
Some Linux distributions contain this package: Debian Sarge (<code>nfs-user-server</code>), OpenSUSE 10.0 (<code>nfs-server</code>).
 
For other distributions you can download sources (for example from Debian repository) and compile it.
 
For other distributions you can download sources (for example from Debian repository) and compile it.
There is a small trick you have to know about runing <code>mountd</code> and <code>nfsd</code> (these two daemons and <code>portmap</code> constitute a user-space server).  You should run them with <code>-r</code> option:
+
There is a small trick you have to know about running <code>mountd</code> and <code>nfsd</code> (these two daemons and <code>portmap</code> constitute a user-space server).  You should run them with the <code>-r</code> option:
 
<pre>
 
<pre>
 
# portmap
 
# portmap
Line 29: Line 29:
 
# rpc.nfsd -r
 
# rpc.nfsd -r
 
</pre>
 
</pre>
The reason is that these daemons check major number of device where the directory to export resides.
+
The reason is that these daemons check the major number of the device where the directory to export resides.
If major equals 0 then daemons suppose that it is NFS and don't want to re-export it. Symptoms are
+
If major equals 0 then daemons assume that it is NFS and don't want to re-export it. Symptoms are
that clients will always get a "permission denied" error. Simfs (the file system on what VE is located)
+
that clients will always get a "permission denied" error. Simfs (the file system on which VE is located)
is assosiated with so called unnamed device, which major equals 0. So, to prevent daemons from checking for
+
is associated with so called unnamed device, in which major equals 0. So, to prevent daemons from checking for
 
re-exporting — just use this <code>-r</code> option.
 
re-exporting — just use this <code>-r</code> option.
  

Revision as of 02:00, 27 October 2007

There are two ways to setup NFS server on common HN: use a user-space NFS server daemon or use an in-kernel implementation of NFS server. Some peculiarities appear if you intend to run NFS server in VE.

Yellowpin.svg Note: for information about NFS client inside VE, see NFS.

Kernel NFS server

Binary RPMs that are provided by OpenVZ community contain kernels compiled without NFS server support. Thus you have to recompile the kernel with CONFIG_NFSD=m. After booting in this kernel you'll be able to use NFS server on HN. In-kernel NFS server runs kernel threads to service requests of clients. But for security reasons kernel threads are prohibited in VEs! So you won't be able to run NFS server inside VE without patching the kernel.

User-space NFS server

Advantage of user-space NFS server is that it can be ran without any support in the kernel. Also if it crashes — there is no crash of the system: just one process crashes, not the kernel! The disadvantage of user-space NFS server is its productivity: no one can be faster than in-kernel implementation.

One well-known implementation of NFS server is "The LINUX User-Space NFS Server" by Olaf Kirch. Some Linux distributions contain this package: Debian Sarge (nfs-user-server), OpenSUSE 10.0 (nfs-server). For other distributions you can download sources (for example from Debian repository) and compile it. There is a small trick you have to know about running mountd and nfsd (these two daemons and portmap constitute a user-space server). You should run them with the -r option:

# portmap
# rpc.mountd -r
# rpc.nfsd -r

The reason is that these daemons check the major number of the device where the directory to export resides. If major equals 0 then daemons assume that it is NFS and don't want to re-export it. Symptoms are that clients will always get a "permission denied" error. Simfs (the file system on which VE is located) is associated with so called unnamed device, in which major equals 0. So, to prevent daemons from checking for re-exporting — just use this -r option.

“The LINUX User-Space NFS Server” by Olaf Kirch implements NFSv2. It means that only files with sizes less than 2GB are processed. If you intend to use such big files then you should use another user-space NFS server implementation: unfs3. It implements v3 of NFS protocol standard.

External links