Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.os.linux.networking > #1585 > unrolled thread
| Started by | Nrth <nrth@invalid.example> |
|---|---|
| First post | 2012-08-15 17:33 +0100 |
| Last post | 2012-08-15 22:31 +0100 |
| Articles | 3 — 2 participants |
Back to article view | Back to comp.os.linux.networking
NFS Share Not Mounting On Debian Client From Centos NFS Server Nrth <nrth@invalid.example> - 2012-08-15 17:33 +0100
Re: NFS Share Not Mounting On Debian Client From Centos NFS Server J G Miller <miller@yoyo.ORG> - 2012-08-15 19:17 +0000
Re: NFS Share Not Mounting On Debian Client From Centos NFS Server Nrth <nrth@invalid.example> - 2012-08-15 22:31 +0100
| From | Nrth <nrth@invalid.example> |
|---|---|
| Date | 2012-08-15 17:33 +0100 |
| Subject | NFS Share Not Mounting On Debian Client From Centos NFS Server |
| Message-ID | <6jvsf9-5n7.ln1@nrth.org> |
Hello, I appear to be having some problems with what many of you may consider to be a fairly trivial matter (I know I did at first). I have a Centos 6.3 NFS server that happily hands out shares to other Centos 6.3 NFS clients but due to some configuration mistake I've made I can't mount the same (one and only) share on a Debian Squeeze NFS client. I would be most grateful if any kind soul/s would put an end to further hair loss and wood-for-the-trees-ness that has consumed much of my afternoon today. As an aside, I'm an X-Files fan which is just about to become apparent to many. On the Centos 6.3 x86_64 server in /etc/exports : /home/pete/ doggett.nrth.lab(rw,no_root_squash) The NFS-related services running on this server all appear in order in that I have rpc.statd, rpc.mountd and rpcbind all correct and present. On the Debian Squeeze amd64 client I have rpcbind and rpc.statd running. I've tried modifying /etc/default/nfs-common on the Debian client to include the following line : NEED_IDMAPD=yes .. as I believe the NFS server on the Centos box is running that version. (As it turns out, it doesn't matter whether I have that option enabled or not, I still cannot browse to the mounted share using Nautilus.) On the Debian Squeeze amd64 client I run this command : mount -t nfs tooms.nrth.lab:/home/pete /home/pete/serverdocs There is no error shown and I get an HD icon on my desktop, but if I try to navigate into this share using Nautilus, I receive an access denied message stating possible reasons for this error as lack of permissions. Fair enough, but I can't see what to change. Despite the above error, I can access the mounted share as root on the Debian client and modify files. However, I cannot even change into that directory as user 'pete' on the Debian client. The only possible cause I can think of is UIDs that are different and not being mapped. For example, on the server the user 'pete' has a UID and GID of 500 whereas on the client the same account has 1000. Any help or information would be gratefully received. Regards, Pete.
[toc] | [next] | [standalone]
| From | J G Miller <miller@yoyo.ORG> |
|---|---|
| Date | 2012-08-15 19:17 +0000 |
| Message-ID | <k0gsjv$144$2@dont-email.me> |
| In reply to | #1585 |
On Wednesday, August 15th, 2012, at 17:33:10h +0100, Nrth explained:
> I have a Centos 6.3 NFS server that happily hands out shares to
> other Centos 6.3 NFS clients
You have failed to state which version of NFS you are running
on the Centos systems and the Debian system.
Is it NFS v3 or v4?
> I've tried modifying /etc/default/nfs-common on the Debian client to
> include the following line :
>
> NEED_IDMAPD=yes
You only need IDMAPD for NFS v4.
If you are running NFS v3 then you need rpc.stad running.
> .. as I believe the NFS server on the Centos box is running that
> version.
Which version?
If you are running NFS v4 with idmapd, then you should also
examine the configuration file /etc/idmapd.conf
Also take a look at
<https://help.ubuntu.com/community/NFSv4Howto>
because the way that directories are exported with NFSv4 is done
in a hierarchy in /etc/exports rather different to the way
NFSv3 is configure.
[toc] | [prev] | [next] | [standalone]
| From | Nrth <nrth@invalid.example> |
|---|---|
| Date | 2012-08-15 22:31 +0100 |
| Message-ID | <b2htf9-ds3.ln1@nrth.org> |
| In reply to | #1586 |
On Wed, 15 Aug 2012 19:17:20 +0000, J G Miller <miller@yoyo.ORG> wrote: > On Wednesday, August 15th, 2012, at 17:33:10h +0100, Nrth explained: > >> I have a Centos 6.3 NFS server that happily hands out shares to other >> Centos 6.3 NFS clients > > You have failed to state which version of NFS you are running on the > Centos systems and the Debian system. > > Is it NFS v3 or v4? Hello, Thanks for the reply and sorry for various omissions. I believe the server is NFSv4. The (Debian) client I'm not so sure about as it appears to be able to talk all three versions. By the way, there are no firewalls active between these two machines. After checking your supplied URL and digging a little deeper, I found with which I'm happy although it may not be to everyone's taste. First of all, the /etc/exports line as it stands now. /home/pete/ doggett.nrth.lab(rw) I've removed 'no_root_squash' and indeed root on the client cannot access the mounted NFS share, neither can any other user which was as expected so that'll do for now. Any other user that is, apart from myself. I found this very useful page and within was the solution which I found most suitable. http://www.troubleshooters.com/linux/ nfs.htm#_Mounting_an_NFS_Share_on_a_Client The solution (for me) was to change the UID and GID of my user account on the Debian box to match that of the server's record of the same username. So 'pete' on both boxes is now '500' in /etc/passwd and /etc/group. A quick chown -R pete:pete /home/pete and a (reflex) reboot and the following mount command worked as before but this time I was able to access the share. mount -t nfs4 tooms.nrth.lab:/home/pete /home/pete/serverdocs Thanks again for the reply and for the motivation. Regards, Pete.
[toc] | [prev] | [standalone]
Back to top | Article view | comp.os.linux.networking
csiph-web