Multiple NFS Servers (Gluster NFS in 3.x, unfsd, knfsd, etc.)
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
- To: 'Gluster Devel' <gluster-devel@xxxxxxxxxx>
- Subject: Multiple NFS Servers (Gluster NFS in 3.x, unfsd, knfsd, etc.)
- From: Gordan Bobic <gordan@xxxxxxxxxx>
- Date: Tue, 05 Jan 2010 10:33:47 +0000
- User-agent: Thunderbird 2.0.0.22 (X11/20090625)
I'm not sure if this is the right place to ask, but Google has failed me
and it is rather related to the upcoming Gluster NFS export feature.
What I'm interested in knowing is whether it will be possible to run
Gluster NFS export for glfs mounts while using knfsd as per standard for
exporting non glfs paths?
e.g. if /home is a glfs mounted volume and /usr/src is on a raw block
device, will it be possible to have /home handled by the glfs NFS export
while having /usr/src handled by the native knfsd?
As far as I can tell main problem with this setup is that I can't see
how will RPC mountd on the server would know which NFS server should be
handling the connection, if there were more than one (could more than
one even be registered with RPC/portmap?).
One possible work-around I can think of is to have one daemon listen for
NFS connections on TCP and the other on UDP, but this is a bit lame. The
only other way I can think of doing it is to put something like an
OpenVZ VM (jail/zone/chroot) on the server, bind mount one lot of
volumes that needs exporting onto that, and then export each lot from a
separate machine (one virtual one physical), from different IPs. Having
said that, is it possible to get portmap/mountd to bind a different NFS
daemon to different IPs without the need for virtualization?
Gordan
[Index of Archives]
[Gluster Users]
[Ceph Users]
[Linux ARM Kernel]
[Linux ARM]
[Linux Omap]
[Fedora ARM]
[IETF Annouce]
[Security]
[Bugtraq]
[Linux]
[Linux OMAP]
[Linux MIPS]
[eCos]
[Asterisk Internet PBX]
[Linux API]