Tiziana Manfroni | 24 Jul 14:29 2012
Picon

NFS no traslate UID,GID

Hi, I haxe a NFS server with NIS on RHEL6.3. When the client mounts a directory 
the system doesn't traslate UID e GID but uses nobody.

On server
The /etc/exports :
/users                       192.168.114.101(rw,sync)

/etc/default/nfs-common
NEED_IDMAPD=yes

/etc/idmapd.conf
Domain = domainserver
Nobody-User = nfsnobody
Nobody-Group = nfsnobody
Method = nsswitch

/etc/nsswitch
passwd:     files nis
group:      files nis

The file /etc/fstab on client is
serverNFS:/users    /users   nfs4  rw,bg,intr,soft,sec=sys 0 0

but ls -la /users
drwx------  28 nobody nobody 4096 23 lug 17:24 utente

How can I resolve this problem?

Thanks
 	Tiziana
(Continue reading)

Ian Mortimer | 25 Jul 04:09 2012
Picon
Picon

Re: NFS no traslate UID,GID

On Tue, 2012-07-24 at 14:29 +0200, Tiziana Manfroni wrote:

> Hi, I haxe a NFS server with NIS on RHEL6.3. When the client mounts a directory 
> the system doesn't traslate UID e GID but uses nobody.

Is rpcidmapd running?  

If it is:

   umount _the_nfs_share_
   service rpcidmapd restart
   mount _the_nfs_share_

If it's not running:

   umount _the_nfs_share_
   service rpcidmapd start
   chkconfig rpcidmapd on
   mount _the_nfs_share_
   
--

-- 
Ian

Feldt, Andrew N. | 25 Jul 15:39 2012
Picon

Re: NFS no traslate UID,GID

On Jul 24, 2012, at 9:09 PM, Ian Mortimer wrote:

> On Tue, 2012-07-24 at 14:29 +0200, Tiziana Manfroni wrote:
> 
>> Hi, I haxe a NFS server with NIS on RHEL6.3. When the client mounts a directory 
>> the system doesn't traslate UID e GID but uses nobody.
> 
> Is rpcidmapd running?  
> 
> If it is:
> 
>   umount _the_nfs_share_
>   service rpcidmapd restart
>   mount _the_nfs_share_
> 
> If it's not running:
> 
>   umount _the_nfs_share_
>   service rpcidmapd start
>   chkconfig rpcidmapd on
>   mount _the_nfs_share_
> 
> -- 
> Ian

Note that this is due to the fact that NFS mounts are done before ypbind is
started.  And, rpcidmapd now caches its results, so the translation to
'nobody' remains even after ypbind is started.  I consider this a bug that
will have to be addressed (maybe by starting ypbind before NFS mounts?)
so that NFSv4 mounts can truly function in an NIS environment.
(Continue reading)

Hearn, Stan J. | 27 Jul 15:48 2012

How do you know when a reboot is required after yum update?

My Package Updater (PUP) in Centos 5 somehow knows when a reboot is needed after an update.

How does it know?

My concern is that we have systems that appear in satellite to be up-to-date.  But if they were never rebooted
after an update, they really aren't.

I see this question asked frequently, however I've not found the answer.

Thanks,
Stan

Greg Swift | 27 Jul 15:58 2012
Picon

Re: How do you know when a reboot is required after yum update?

On Fri, Jul 27, 2012 at 8:48 AM, Hearn, Stan J. <stan.hearn@...> wrote:
> My Package Updater (PUP) in Centos 5 somehow knows when a reboot is needed after an update.
>
> How does it know?
>
> My concern is that we have systems that appear in satellite to be up-to-date.  But if they were never
rebooted after an update, they really aren't.
>
> I see this question asked frequently, however I've not found the answer.

I recently read some discussion on this....

http://fedoraproject.org/wiki/Features/OfflineSystemUpdates
http://fedoraproject.org/wiki/Talk:Features/OfflineSystemUpdates


Gmane