Incremental Index-Backup

thanks bingo...

hmmm.. it does not mention level 1..  ??

Thanks so far for all the help.

I'll try the suggestions and will post it here.

Just check at pool , level is not selected apart from Index and Bootstrap a save sets.

Hi Wallce,

I tried both of your commands:

savegrp -O -pv group-name

savegrp -O -l 9  -pv group-name

Both bring back all the clients in the group and it is stated that a full backup of the index will be performed:

What I find odd is that the pool which is mentioned in the output is not the INDEX pool. It says that the index will be written in the data pool.

EDIT: I fixed the issue with the wrong shown pool. I removed the space character between ":" and "*" in the save set field.

Don't mix up the client backup with the client index backup. Your client backup seems to use the correct level but the index backup doesn't. The reason for this is name resolution inconsistencies.

Do you have the same problem with other client index backups? If so, it's a name resolution issue with your Networker server client and not this client.

If not, it's related to this particular client name and we should continue troubleshooting this client.

Please show us the result of nslookup of the client name and ip-address

Check the name reported by the client itself. Run "nsradmin -p nsrexec -s client_name" and then type "p nsrla". What is the name reported here? Is it consistent with the name you use for the client resource?

Show us the content of the index directory for this client .../nsr/index/client_name

Hi MagnusR,

the daemon.log states this during the incr backup:

nw-server nsrd NSR info Savegroup Info: Unable to find any full backups of the save set 'client:index' in the media database. Recommending that a full backup be performed.

nw-server nsrd NSR info nw-server:index:client saving to pool 'INDEX' (INDEX)

nw-server nsrd NSR info nw-server:index:client done saving to pool 'INDEX' (INDEX) 297 MB

Ok, if this happens for all index backups then the problem is with the Networker server itself.

Are all NW server backups promoted to full or only the index backups?

Hi MagnusR,

this is the behaviour for all of my clients.

So nslookup for the client:

forward ok

backward ok

Nslookup for NetWorker server:

forward ok

backward ok

nsradmin -p nsrexec -s client

NetWorker administration program.

Use the "help" command for help, "visual" for full-screen mode.

nsradmin> p nsrla

                        type: NSRLA;

                        name: client.fqdn;

              nsrmmd version: ;

             nsrsnmd version: ;

NW instance info operations: ;

       NW instance info file: ;

          installed products: ;

                auth methods: "0.0.0.0/0,nsrauth/oldauth";

           max auth attempts: 8;

               administrator: "isroot,host=nw-server.old",

                              "isroot,host=nw-server.fqdn.old",

                              "isroot,host=nw-server",

                              "isroot,host=nw-server.fqdn",

                              "user=root,host=client.fqdn",

                              "user=root,host=localhost";

                        arch: x86_64;

                 kernel arch: x86_64;

                    CPU type: x86_64;

                machine type: desktop;

                          OS: Linux 3.0.101-0.47.99-default;

           NetWorker version: 8.2.3.0.Build.1100;

              client OS type: Linux;

                        CPUs: 4;

                     MB used: 188136;

                  IP address: 10.4.9.143, 10.4.9.145;

  environment variable names: ;

The name reported as client name is the FQDN. In the NMC the client ressource is without the FQDN. Under /nsr/index is also the hostname without the FQDN stated.

ll /nsr/index/client

total 8.0K

-rw-r--r--  1 root root  970 Aug  9  2016 README

drwx------ 37 root root 4.0K Dec 18 11:02 db6/

-rw-------  1 root root    0 Aug  9  2016 v6ck.lck

Hi Prince,

nothing selected here.

Since all client index backups are affected, this could be related to client id issues.

Please review the following kb: 

NetWorker: How to use nsrclientfix, error "Cannot find attribute 'client id'"

https://support.emc.com/kb/333478

nsrclientfix man page.pdf  

https://emcservice.my.salesforce.com/servlet/fileField?retURL=https%3A%2F%2Femcservice--c.na55.visual.force.com%2Fapex%2…

For now, please run step 1-3 under resolutions.  Then take a look at the output file.  Look for any reference to the backup server.

Only the index backups. Running incr backups from any machine works fine.

May I suggest the following:

  -  move the file index of one specific client

  -  run an incr backup of this client (NW should run a full index backup)

  -  re-run an incr backup. The index backup should now come up with level 9.

If this works, your file/media index is defective. Otherwise it is NetWorker itself causing the problem.

I did steps 1 to 3 from the kb and nsrclientfix did find some ID issues. But our client was not in the list. So I suggest there is no ID problem here.

Hi Bingo,

I did exactly what you suggested and the index backup is still a full .

I think it is time to open a SR with DELL EMC now.

Thanks anyone for the help so far. I will let you know what the support brings up.