Failures with: server_export -option nfsv4only

I already did that as shown in the first few lines of my posting.  And, yes, I did triple check that am referencing the correct server_2 and slot_2.

I even verified the inodes for the netd file were the same for server_2 and slot_2.  I did edit what I believe is the correct file and rebooted the correct datamover. I also rebooted the whole appliance just incase....

Hi,

Ensure that the /nas/server/slot_ /config file where is the slot number   of the Data Mover has the hivers set to 4. The nfs entry in the /nas/server/slot_ /config file should appear similar to :

nfs config hivers=4

Then, reboot the Data Mover.

It would be /nas/server/slot_2/config

PLEASE NOTE:

There was a typo with the accesspolicy option in the original server_mount command:

server_mount server_2 -option accesspolity=MIXED,nooplock,rwlock,nfsv4delegation=NONE queue /queue

I have done a permenent  unmount of "/queue" and repeated the commands with the correct syntax.

The nfsv4only export option continues to fail with the same error in my original post:

server_export server_2 -Protocol nfs -option access=10.0.13.0/255.255.252.0,root=10.0.13.0/255.255.252.0,nfsv4only -comment 'NFS v4 only export for

queue' /queue

server_2 :

Error 22: server_2 : Invalid argument

Export Error: Nfs has been started with HI_VERS 3. Cannot export as nfsv4 only.

Did you check the config file instead the netd ?

Thanks,

Content of /nas/server/slot_2/netd

routed

nfs start hivers=4

statd

lockd

pax

rquotad action=start

ndmp port=10000

xattrp

snmpd

kerbinit

Content of /nas/server/slot_2/config

nfs config

I also find that after editing /etc/server/slot_2/netd, setting hivers=4, then rebooting the datamover,  nfs v4 is not running.

Here are some specifics showing the specific release, content of my netd, the reboot operation and the NFS failed status after the reboot.

Suggestions???

[nasadmin@Celerra-v6 ~]$ cat /etc/emc-release

EMC Celerra Control Station Linux release 3.0 (NAS 6.0.36)

[nasadmin@Celerra-v6 ~]$ cat /nas/server/slot_2/netd

routed

nfs start hivers=4

statd

lockd

pax

rquotad action=start

ndmp port=10000

xattrp

snmpd

kerbinit

[nasadmin@Celerra-v6 ~]$ server_cpu server_2 -reboot -monitor now

server_2 : reboot in progress 5.0.0.0.0.0.0.3.3.3.3.3.3.3.4.done

[nasadmin@Celerra-v6 ~]$ server_nfs server_2 -v4

server_2 :

NFSv4 is not enabled

Restart system and use nfs option hivers=4

[nasadmin@Celerra-v6 ~]$ server_nfs server_2 -v4 -service -start

server_2 :

Error 4020: server_2 : failed to complete command

File to edit is

/nas/server/slot_2/config

And put there hivers=4

And you get nfsv4 started

--

Jussi

Thanks.

I was using V6 documentation and the section on the NFS manual, page 41 "Specify NFS v4 access" said to edit the start command in the netd file.

I misread the reference on page 28 that said "the /nas/server/slot_x/conf file".  I didn't take it literally and missed it since the section that explictly talked about "netv4only" explicted said the start command in netd.

It was a little obscure becuase the "hivers=4" I put into /net/server/slot_2/netd was not rejected or reported nay warnings on the console.

I never knew nfs v4 failed to start until long after trying to figure out what the nfsv4only error really meant.

BTW: 

I was very unhappy that the ysstem let me put an invalid parameter into netd then nicely ignore it. Just for fun I put a bogus parameter into /nas/server/slot_2/netd the reboot went into limbo.....  I never got an error, I never got the "reboot in progress"...

This was ignored:        nfs start hivers=4

This caused limbo:      nfs start thisIsStupid

Yes, this was eventually user error but I stared at it too long and missed it.  Also, the CD I downloaded did not have release notes.  Where was the drastic change between the v5 and v6 configuration described?

Again, thanks.