ndmp error

Check admin guide and release supplement - both have this covered pretty well and will give you enough to start. You did specify ndmp user (or root) & password for the client?

I've gone through everything that the admin guide has. Everything appears to conform to the NDMP setup but whether I use All or list all the individual save set mounts, I get a connection refused error when I try to start the backup. Part of the problem is that all of the information I can find on the Celerra is only related to using the Celerra as the backup devide, not backing up the data on the Celerra. Is there some Celerra related documentation that I need to look at?

Message was edited by:
Michael Fay

You can't use saveset All with Celerra - that's also in admin guide.

Let's start from connection itself which according to error seems to be the problem - what username do you use? Actually, can you list from nsradmin client resource for this client and post it here?

type: NSR client;
name: 137.159.9.78;
server: networker;
comment: NDMP for NAS;
Save operations: ;
archive services: Disabled;
schedule: Default;
browse policy: 45 days;
retention policy: 45 days;
directive: ;
group: NDMP;
save set: /cifs01, /nfs01, /nfsbragifsoraback,
/nfsbragipaoraback, /nfsc2cdev, /nfsc2cprd,
/nfsc2cdmo, /nfsc2cstg, /nfsc2ctrn, /nfsc2ctst;
remote access: "user=root,host=networker";
remote user: ndmpuser;
password: ********;
backup command: nsrndmp_save -T dump;
application information: HIST=y, UPDATE=y, DIRECT=y;
ndmp: Yes;
aliases: 137.159.9.78;
archive users: ;
storage nodes: nsrserverhost;
clone storage nodes: ;
client OS type: ;
CPUs: ;
NetWorker version: ;
enabler in use: No;
licensed applications: ;
licensed PSPs: ;

Message was edited by:
Michael Fay

I'm not sure that alias is allowed to be IP address. Same applies to the name of the client resource. It should point to the name of data mover.

Also, saveset usually points to certain data mover within celerra and then path (eg. /root_dmX/something) - at least from what I have seen and worked with.

Okay, I remade the client: This points directly to the datamover:


hostname: griffin.pepperdine.edu;
nsradmin> print name:griffin.pepperdine.edu
type: NSR client;
name: griffin.pepperdine.edu;
server: networker;
comment: "NAS ";
Save operations: ;
archive services: Disabled;
schedule: Default;
browse policy: Month;
retention policy: Year;
directive: ;
group: NDMP;
save set: /griffin/nfs01, /griffin/nfsbragifsoraback,
/griffin/nfsbragipaoraback, /griffin/nfsc2cdev,
/griffin/nfsc2cdmo, /griffin/nfsc2cprd,
/griffin/nfsc2cstg, /griffin/nfsc2ctst,
/griffin/nfsc2ctrn, /griffin/cifs01;
remote access: "user=root,host=networker";
remote user: ndmpuser;
password: ********;
backup command: nsrndmp_save -T dump;
application information: HIST=y, UPDATE=y, DIRECT=y;
ndmp: Yes;
aliases: griffin, griffin.pepperdine.edu;
archive users: ;
storage nodes: nsrserverhost;
clone storage nodes: ;
client OS type: ;
CPUs: ;
NetWorker version: ;
enabler in use: No;
licensed applications: ;
licensed PSPs: ;


I get the following error when I run the backup.

06/21/06 17:21:56 savegrp: griffin.pepperdine.edu:/griffin/nfsc2cdev will retry 1 more time(s)
griffin.pepperdine.edu:/griffin/nfsbragipaoraback: No full backups of this save set were found in the media database; performing a full backup
* griffin.pepperdine.edu:/griffin/nfsbragipaoraback ******************************************************************
* griffin.pepperdine.edu:/griffin/nfsbragipaoraback ------ E R R O R -------
* griffin.pepperdine.edu:/griffin/nfsbragipaoraback Not able to connect or authorize the client 'griffin.pepperdine.edu'
* griffin.pepperdine.edu:/griffin/nfsbragipaoraback ******************************************************************
griffin.pepperdine.edu: /griffin/nfsbragipaoraback level=full, 0 KB 319704:21:56 0 file
* griffin.pepperdine.edu:/griffin/nfsbragipaoraback backup failed.
* griffin.pepperdine.edu:/griffin/nfsbragipaoraback nsrndmp_save: Diagnostic: Save session closed with NW server sucessfully
06/21/06 17:21:56 savegrp: griffin.pepperdine.edu:/griffin/nfsbragipaoraback will retry 1 more time(s)

I've tried using both the ndmpuser and root users. Any ideas?

Under assumtion eevrything else is configured correctly - no idea (just for the sake of obvious things - is ndmp enabled on Celerra side?). You may wish to involve support on this - they will probably then have a webex to you and solve it in no time.

NDMP is enabled. The server is not firewall blocked (Solaris 9).


Is there a way to test and see if Networker can actually login to the NDMP user on the Celerra? a command line?

I guess you could put both ndmpd on Celerra and nsrd on backup server to debug mode and rerun your test. That should give you more data.

Set the storage node attribute to the name of the Celerra. Make sure you can connect to port 10,000 on the Celerra (basic telnet to that port then disconnect) to verify it is open.

Do you have the correct NDMP licenses installed on your server?

Did you set up your tape devices properly? What is the output of this command:

server_devconfig servername -l -s -n

What Donovan says (devices) make sense only if you are using Celerra as storage node. Otherwise you can use bck server or any other storage node to direct your saves to.

I hacked it all out with EMC support, and it turns out that 7.1.4 can only support NDMP to a direct attached storage device. So I have to upgrade to at least 7.2 to get DSA support to be able to do what I want to do.

So, I guess I'll upgrade to 7.3 (Scary thought, that).

Yes, that's pretty much best feature in 7.3 I would say. In such case you should go for 7.3.1 and check what available patches are there and see if any of those might be required for your environment.

As a practical matter, nothing.We have other considerations that make going to 7.3 desirable, such as the need to seperate clone retentions from the retentions of the original save set.