[opensuse] nfs issue
![](https://seccdn.libravatar.org/avatar/898103d5a7c5ab5de5d7101122d14af9.jpg?s=120&d=mm&r=g)
Hi all, Yesterday i tried to export from a 12.1 system towards some 11.4 systems (nfs4) and i came across something unexpected (for me at least) In the tree that i export, i have mounted some logical volumes At the receiving end, i see the imported tree, but everything that is mounted (on the nfs server) does not appear. I know that it is nit good practice to re-export nfs-mount, but in this case i try to export a bumch of loccally mounted volumes. Sure i can nfs-export each volume individually, but i hoped i god do this in a more practical way. been googling about "submounts" but no clue yet if either i am doing something wrong, or that it is not possible in the first place. Anyone suggestions? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
![](https://seccdn.libravatar.org/avatar/247f3737bfdd07c80a5411399e9a504c.jpg?s=120&d=mm&r=g)
Hans Witvliet wrote:
Hi all,
Yesterday i tried to export from a 12.1 system towards some 11.4 systems (nfs4) and i came across something unexpected (for me at least)
[snip]
Anyone suggestions?
I don't know much about NFS 4, but I have two suggestions. (1) You might try https://help.ubuntu.com/community/NFSv4Howto (2) you didn't post any information about your config or what you have tried, so there's probably not a lot people can do to help. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
![](https://seccdn.libravatar.org/avatar/f15c560a0ca9454d9c0122ea1a7a54e2.jpg?s=120&d=mm&r=g)
On 22/06/12 09:11, Hans Witvliet wrote:
In the tree that i export, i have mounted some logical volumes At the receiving end, i see the imported tree, but everything that is mounted (on the nfs server) does not appear.
I know that it is nit good practice to re-export nfs-mount, but in this case i try to export a bumch of loccally mounted volumes.
Hi Hans How are your logical volumes mounted? Exporting bind mounted shares fsid0 used to be the recommended way for nfs4 but that has changed: http://wiki.linux-nfs.org/wiki/index.php/Nfsv4_configuration#Exporting_direc... (I think the url has wrapped) You simply export them from wherever they are in the underlying filesystem just as you did with other nfs versions. L x -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
![](https://seccdn.libravatar.org/avatar/898103d5a7c5ab5de5d7101122d14af9.jpg?s=120&d=mm&r=g)
On Fri, 2012-06-22 at 13:09 +0200, lynn wrote:
On 22/06/12 09:11, Hans Witvliet wrote:
In the tree that i export, i have mounted some logical volumes At the receiving end, i see the imported tree, but everything that is mounted (on the nfs server) does not appear.
I know that it is nit good practice to re-export nfs-mount, but in this case i try to export a bumch of loccally mounted volumes.
Hi Hans
How are your logical volumes mounted?
Exporting bind mounted shares fsid0 used to be the recommended way for nfs4 but that has changed:
http://wiki.linux-nfs.org/wiki/index.php/Nfsv4_configuration#Exporting_direc...
(I think the url has wrapped)
You simply export them from wherever they are in the underlying filesystem just as you did with other nfs versions.
tnx Lynn, Dave, all... Perhaps i as indeed a bit too brief. Often i include too much info that might not be relevant. Actually i'm not so sure if it is an nfs4-issue. Some mails on fedora suggest that such behaviour comes with kernel-based nfs-servers, while userland nfs-servers would not suffer from it. Can not tell if that makes any sense or is just an urban legend. Anyway, i did no do anything fancy. On the nfs server my /etc/exports is: (sorry for the line wrap) /local/software/obs 192.168.0.0/24(rw,no_root_squash,no_subtree_check) While on that machine, my fstab is: # /dev/shared/sw_obs /local/software/obs/ btrfs defaults 1 2 # /dev/shared/sw_obs_server_colon /local/software/obs/server: btrfs defaults 1 2 /dev/shared/sw_obs_network_colon /local/software/obs/network: btrfs defaults 1 2 1) i got an mount-point called "/local/software/obs/" That is a 25GB LV 2) i got an mount-point called "/local/software/obs/server:" That got an 65GB lv 3) i got an mount-point called "/local/software/obs/network:" That got an 40GB lv What i initially did, was mounting LV-1 and afterwards mounting LV-2 and LV-3 On the nfs-server itself everything looked as expected... On an other machine i made the mount point and did manually: mount 192.168.0.2:/local/software/obs /local/software/obs The mount command showed: [long line] 192.168.0.2:/local/software/obs/ on /local/software/obs type nfs4 (rw,relatime,vers=4,rsize=524288,wsize=524288,namlen=255,hard,proto=tcp,port=0,timeo=600,retrans=2,sec=sys,clientaddr=192.168.0.25,minorversion=0,local_lock=none,addr=192.168.0.2) I could see both sub-mount-points (server:/ and network:/) but could not reach the storage that i was expecting to see there. ----------------------------- So what i did next, was exporting & importing the "server:" and "network" directory explicitly. Although that works, it just proves that there is nothing wrong with the LV's. Had a quick look at both mentioned URL's, but they are mostly about kerberos and so-on (which i do't use at home) I fear my problem/misconception is more "fundamental" Hans -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
![](https://seccdn.libravatar.org/avatar/77cb4da5f72bc176182dcc33f03a18f3.jpg?s=120&d=mm&r=g)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 2012-06-23 00:30, Hans Witvliet wrote:
I fear my problem/misconception is more "fundamental"
Have a look at "nohide" in man exports, may be related. - -- Cheers / Saludos, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk/k/UsACgkQIvFNjefEBxotdACg2Rx1nDU0uSzJuaOIpxjeWI9j 4FoAn3/QyPb/hv2ZzPkjjOMO+oo6/wlD =LJRs -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
![](https://seccdn.libravatar.org/avatar/f15c560a0ca9454d9c0122ea1a7a54e2.jpg?s=120&d=mm&r=g)
On 06/23/2012 01:18 AM, Carlos E. R. wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 2012-06-23 00:30, Hans Witvliet wrote:
I fear my problem/misconception is more "fundamental" Have a look at "nohide" in man exports, may be related.
Hi I wonder whether you really do have nfs4 at both ends. How about trying manually: mount -t nfs4 server:/local/software/obs /mnt then have a poke around in /mnt If no, try it with -overs=3 and see if there is any difference. mount -t nfs server:/local/software/obs /mnt -overs=3 Next, have a look to make sure rpc.idmapd is running on both client and server. Check the output of rpc.idmapd -fvvv Easy way is to just check the 'use nfs v4 box' in yast nfs server and yast nfs client. For Carlos' nohide method you have to create a pseudoroot: mkdir /exports mkdir /exports/obs mount --bind /local/software/obs /exports/obs etc/exports now has to look like this: /export 192.168.0.0/24(rw,fsid=0,no_subtree_check) /export/obs 192.168.0.0/24(rw,nohide,no_subtree_check) or use crossmnt: /export 192.168.0.0/24(rw,fsid=0,crossmnt,no_subtree_check) /export/obs 192.168.0.0/24(rw,no_subtree_check) On the client simply mount the root: mount -t nfs4 server:/ /mnt If it works, in fstab: server:/ /mnt nfs4 auto 0 0 Replace /mnt with wherever you want it. HTH L x -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
participants (4)
-
Carlos E. R.
-
Dave Howorth
-
Hans Witvliet
-
lynn