On Fri, Mar 23, 2012 at 12:19 AM, Tom H
<tomh0665@gmail.com> wrote:
I've lost track whether it's when you're mounting the serverA or
serverB export that you're having the nobody problem but do you have
Okay maybe we leave that, i was trying to show I had the same setup on both side, but one was working and 1 wasn't
the same problem when mounting that export from another box? Do you
have the problem when mounting via hostname and not via ip address?
i usually mount via hostname
Are all your "Domain" values the same in all your boxes'
"/etc/idmapd.conf"? Do they all have "Domain = abc.com.au" like the
one that you posted earlier?
yep both idmapd.conf are meant to be the same
>> Do "/var/log/messages" and a verbose mount give you any information on
>> the failure?
no nothing
>>
> so i tried a mount -v ? is that what you meant by verbose, the only thing I
> got was
> Mar 20 13:37:27 max rpc.idmapd[19081]: nss_getpwnam: name 'nobody' does not
> map into domain '
samad.com.au'
I will try that thanks
> got me thinking my nsswitch and some other libraries are not update on
> server B this is the one serving up the bad mount
Is "/etc/nsswitch.conf" the same on your two boxes? Does "getent
hosts" list all of your hosts and their ip addresses? Can you query
LDAP for hostnames on all your boxes?
yep and getent passwd and getent shadow and getent group are the same !
>> (What's the "localdoman" variable in :/etc/idmapd.conf" for?)
>
> don't know !
I've never seen "localdomain/localdoman" as an "/etc/idmapd.conf" stanza.
okay, I think it was in there as part of the default
I have to try and get the box to boot now, after the update to the kernal and system files its not boot ;(
[snip]