NFS mount -

Bob Goodwin - Zuni, Virginia, USA bobgoodwin at wildblue.net
Fri Nov 14 23:10:10 UTC 2014


On 11/14/14 16:32, Rick Stevens wrote:
> On 11/14/2014 12:12 PM, Bob Goodwin - Zuni, Virginia, USA issued this
> missive:
>
Sorry, that was an error, I picked the wrong command from history, it 
should have been:

[root at box10 bobg]# mount 192.168.1.48:/nfs4exports/data  /mnt/BOX48
mount.nfs: Connection timed out

With my poor vision I miss stuff like that easily.
> You have several problems. First, the NFS server is set up to export
>
> 	/nfs4exports
> 	/nfs4exports/data
> 	/nfs4exports/home
>
> It is NOT exporting /mnt/nasdata.
/mnt/nasdata was an artifact of the Freenas server I am replacing.
> You can try "showmount -e" on the
> NFS server to see what it's actually exporting. I think you'll find the
> things being exported all start with /nfs4exports.
[root at box10 bobg]# showmount -e 192.168.1.48
clnt_create: RPC: Port mapper failure - Unable to receive: errno 113 (No 
route to host)
>   If that's the case,
> try:
>
> 	mount 192.168.1.48:/nfs4exports /mnt/box48
[root at box10 bobg]# mount 192.168.1.48:/nfs4exports /mnt/box48
mount.nfs: Connection timed out

> See my other comments below.
>
>> Does nothing until it eventually times out. I can ssh into the server
>> and see all the files. I tried to configure it to be nearly the same as
>> another NFS server that has been working well.
>>
>> [bobg at box48 ~]$ cat /etc/exports
>> #
>> #    /etc/exports
>>
>> /nfs4exports
>> 192.168.1.0/24(ro,sync,insecure,no_root_squash,no_subtree_check,fsid=0)
>>
>> /nfs4exports/data
>> 192.168.1.0/24(rw,sync,insecure,no_root_squash,no_subtree_check)
>>
>> /nfs4exports/home
>> 192.168.1.0/24(rw,sync,insecure,no_root_squash,no_subtree_check)
>>
Changed as below:

> Looking at your bindmounts, the actual mountpoints that you're using
> aren't absolute. I think you want:
>
> 	/home/data    /nfs4exports/data    none    rw,bind    0 0
> 	/home/home    /nfs4exports/home    none    rw,bind    0 0
/etc/fstab changed as below, but what I used was copied from the other 
working NFS:

# bind mounts

/home/data    /nfs4exports/data    none    rw,bind    0 0

/home/home    /nfs4exports/home    none    rw,bind    0 0
>
> And ensure that the "/nfs4exports", "/nfs4exports/data" and
> "/nfs4exports/home" directories all exist BEFORE you execute the
> bindmount.

ls / shows drwxr-xr-x.   4 root root   28 Nov 14 13:00 nfs4exports

and
[bobg at box48 ~]$ ls -l  /nfs4exports
total 0
drwxr-xr-x. 2 root root 6 Nov 14 13:00 data
drwxr-xr-x. 2 root root 6 Nov 14 13:00 home



/ "BEFORE you execute the bindmount" I'm not sure how to do/verify that? /

>> I have been referring to the Fedora Project NFS guide but apparently I
>> am missing something. I tried systemctl stop iptables on the server, no
>> change. I routinely mount the other NFS so I assume the problem is not
>> in "Firewalld."
>>
>> One difference is that the server in question has two identical drives
>> configured Raid1. Apparently XFS is preferred or required, I'm nor sure
>> but Anaconda kept changing ext4 to xfs so I assumed they know better
>> than I do?
> ----------------------------------------------------------------------
> - Rick Stevens, Systems Engineer, AllDigital    ricks at alldigital.com -
> - AIM/Skype: therps2        ICQ: 22643734            Yahoo: origrps2 -
> -                                                                    -
> -   Which is worse: ignorance or apathy?  I don't know.  Who cares?  -
> ----------------------------------------------------------------------
Thanks,

Bob

-- 
http://www.qrz.com/db/W2BOD
box10  Fedora-20/64bit Linux/XFCE



More information about the users mailing list