Media Server cannot communicate with Data Domain with DDBOOST after DR

Question

Hi all,

We have NetBackup (7.5.0.4) using the Symantec VCS GCO which is a wide area network cluster on master server. . We are using tape drive and DD (DD670) with DDBOOST as the storage. The system all running fine during production for tape backup and DD backup.

During the DR (test) the media server cannot communicate with the DD anymore. Since this is a wide area network cluster the master server will change to the one on the DR site and EMMSERVER name will be the same but the IP will be changed (we have to change it on the DNS). All the tape backup also running fine only the DD backup all failed.

I went to the NetBackup GUI tried to do the "Media & Device Management -> Credentials -> Storage Servers -> right click to our dd -> Update Storage Server Details" it will give me this error:

"Failed to update the following Storage Server: ddroc1bck invalid command parameter(20)

RDSM has encountered an issue with STS where the server was not found: getStorageServerInfo"

So what went wrong? Any log I can check? Any solution.

Thank you.

Regards,

Iwan Tamimi

ble1 · Answer

Is it possible that hosts records still show old IP?

iwantam · Answer

Hi Hrvoje Crevlin, How to check it? On the Data Domain I changed the master server IP manually.

ble1 · Answer

If you changed it there, then I guess it is changed.  It seems as if something is not yet changed (or cached).  You may wish to take advantage of current DD DR session going at Ask the Expert: Using Data Domain Storage in your Disaster Recovery Plan and explain in detail what you had, what you did and what was the outcome.  Though that session is not intended to do technical troubleshooting, from the steps made they might be able to spot what has been missed or what needs to be changed/checked.

iwantam · Answer

Hi Hrvoje Crvelin ,

Thanks for the suggestion I will look at that site.

I have one more finding, it could be the problem, but I still don't know how to fix it.

Our NetBackup Master Server for example the name of the cluster service is NBMSBCK (which is also the name of of the EMMSERVER) during the normal production it will reside on nbmspd01 server (hostname) during the DR it will reside on nbmsdr01 (hostname). And this is our finding:

on the Data Domain on the production and DR site:

During the normal Production

dd01bck# ddboost show connections

Active Clients: 2

Clients:
Client     Idle   CPUs   Memory(MiB)   Plugin Version                                                             OS Version   Application Version
--------   ----   ----   -----------   --------------   --------------------------------------------------------------------   -------------------
XXXbck   NO       24        16,373   2.5.0.3-314845      Microsoft Windows Server 2008 Service Pack 2 (build 6002), 64-bit    NetBackup: 7.5.0.4
nbmspd01 NO        4         7,874   2.5.0.3-314845                                Linux 2.6.32-131.0.15.el6.x86_64 x86_64    NetBackup: 7.5.0.4
YYYY     YES      24        32,768   2.5.0.3-314845   Microsoft Windows Server 2008 R2 Service Pack 1 (build 7601), 64-bit    NetBackup: 7.5.0.4
--------   ----   ----   -----------   --------------   --------------------------------------------------------------------   -------------------

During the DR (test)

dd01bck# ddboost show connections

Active Clients: 2

Clients:
Client     Idle   CPUs   Memory(MiB)   Plugin Version                                                             OS Version   Application Version
--------   ----   ----   -----------   --------------   --------------------------------------------------------------------   -------------------
XXXbck   NO       24        16,373   2.5.0.3-314845      Microsoft Windows Server 2008 Service Pack 2 (build 6002), 64-bit    NetBackup: 7.5.0.4
nbmspd01 NO        4         7,874   2.5.0.3-314845                                Linux 2.6.32-131.0.15.el6.x86_64 x86_64    NetBackup: 7.5.0.4
YYYY     YES      24        32,768   2.5.0.3-314845   Microsoft Windows Server 2008 R2 Service Pack 1 (build 7601), 64-bit    NetBackup: 7.5.0.4
--------   ----   ----   -----------   --------------   --------------------------------------------------------------------   -------------------

As you can see the active connection on normal Production and DR is going to nbmspd01 which is production server. Logically during the DR the connection should go to client nbmsdr01.

Obviously something wrong here. Is there any configuration need to change during the DR?

Regards,

Iwan

iwantam · Answer

Hi Hrvoje Crvelin ,

Thanks for the suggestion I will look at that site.

I have one more finding, it could be the problem, but I still don't know how to fix it.

Our NetBackup Master Server for example the name of the cluster service is NBMSBCK (which is also the name of of the EMMSERVER) during the normal production it will reside on nbmspd01 server (hostname) during the DR it will reside on nbmsdr01 (hostname). And this is our finding:

on the Data Domain on the production and DR site:

During the normal Production

dd01bck# ddboost show connections

Active Clients: 2

Clients:
Client     Idle   CPUs   Memory(MiB)   Plugin Version                                                             OS Version   Application Version
--------   ----   ----   -----------   --------------   --------------------------------------------------------------------   -------------------
XXXbck   NO       24        16,373   2.5.0.3-314845      Microsoft Windows Server 2008 Service Pack 2 (build 6002), 64-bit    NetBackup: 7.5.0.4
nbmspd01 NO        4         7,874   2.5.0.3-314845                                Linux 2.6.32-131.0.15.el6.x86_64 x86_64    NetBackup: 7.5.0.4
YYYY     YES      24        32,768   2.5.0.3-314845   Microsoft Windows Server 2008 R2 Service Pack 1 (build 7601), 64-bit    NetBackup: 7.5.0.4
--------   ----   ----   -----------   --------------   --------------------------------------------------------------------   -------------------

During the DR (test)

dd01bck# ddboost show connections

Active Clients: 2

Clients:
Client     Idle   CPUs   Memory(MiB)   Plugin Version                                                             OS Version   Application Version
--------   ----   ----   -----------   --------------   --------------------------------------------------------------------   -------------------
XXXbck   NO       24        16,373   2.5.0.3-314845      Microsoft Windows Server 2008 Service Pack 2 (build 6002), 64-bit    NetBackup: 7.5.0.4
nbmspd01 NO        4         7,874   2.5.0.3-314845                                Linux 2.6.32-131.0.15.el6.x86_64 x86_64    NetBackup: 7.5.0.4
YYYY     YES      24        32,768   2.5.0.3-314845   Microsoft Windows Server 2008 R2 Service Pack 1 (build 7601), 64-bit    NetBackup: 7.5.0.4
--------   ----   ----   -----------   --------------   --------------------------------------------------------------------   -------------------

As you can see the active connection on normal Production and DR is going to nbmspd01 which is production server. Logically during the DR the connection should go to client nbmsdr01.

Obviously something wrong here. Is there any configuration need to change during the DR?

Regards,

Iwan

ble1 · Answer

I'm not so familiar with NB to be able to give you qualified answer, but it seems odd that it shows active connection to server which should be off. If I would need to guess, I would say nbmsdr01 IP is resolved to nbmspd01 and this is why DD unit is reporting it this way. You did say DNS change was done so could it be that this is the reason?  Eventually, you can check what does traceroute and nslookup do against nbmspd01 from DD unit.

Data Domain

Media Server cannot communicate with Data Domain with DDBOOST after DR

Was this post helpful?