Server Status: Communication with the server has not been initiated or the server status has not been retrieved

[bluedemon]

I install Netbackup 7.6.1 in our new Windows Server 2012 server and configure the backup to initiate oracle RMAN backup in our UNIX server. The RMAN backup fail with the following error:

RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: failure of backup command on ch00 channel at 10/11/2015 19:24:56
ORA-19506: failed to create sequential file, name="bk_66_1_892833531", parms=""
ORA-27028: skgfqcre: sbtbackup returned error
ORA-19511: Error received from media manager layer, error text:
   VxBSACreateObject: Failed with error:
   Server Status:  Communication with the server has not been initiated or the server status has not been retrieved from the server

Attached is the RMAN log error, bprd log from the Windows 2012 server and dbclient log from the Unix Server

Kindly advise what should be done?

[RiaanBadenhorst]

Hello,

Can you post from master server

bptestbpcd -host "oracle server"

From client

bpclntcmd -pn

bpclntcmd -self

Are you using Oracle intelligent policy or rman script?

Have you linked library on the client?

[Elango_G1]

Hi

This is common after upgrading into new version. Have you tried to link the oracle and NetBackup agent?

go to /usr/openv/netbackup/bin

run ./oracle_link (try with oracle user)

Regards

Elango

[bluedemon]

Attached is the result of command from the Master server and the Client (Oracle Server)

The Netbackup library was already linked to the client. Attached is the result of the linking

[bluedemon]

Yes, the library linking between the Oracle and Netbackup agent is already done before I encountered this error

[RiaanBadenhorst]

Hi,

Looks ok. Have you been able to do a backup of the FS of the solaris client?

Are you using rman script or Oracle Intelligent policy?

[bluedemon]

I am using the RMAN script. I can backup the files of the Solaris client

[RiaanBadenhorst]

Can you post bpbrm log too? Does it write and stop or nothing at all?

[Marianne]

Is date and time on the master and client the same? A minute or two is not a problem but the timestamps in dbclient and bprd logs do not correspond.

[bluedemon]

The time differences between the server and client is only 3 minutes. Here I attached the full bprd and dbclient log from yesterday (11 August 2015)

[Marianne]

Before we open the logs - please show us the text in Activity Monitor Details of a failed job.

This will give us timestamps and PIDs to look for in the logs.

[bluedemon]

Attached is the detail for the failed job from Netbackup server

[RiaanBadenhorst]

Hi,

Can you master resolve the client reverse? In the job details its saying the client is 192.168.10.254, not seemglv.

Please check that bpcltncmd -ip 192.168.10.254 returns seemglv.

[bluedemon]

Attached is the result of the command

[Marianne]

starting backup job (jobid=1074) for client 192.168.10.254

Riaan is right - you seem to have an IP address in the policy, not client hostname.
If forward and reverse name lookup is fine, why is that?

NBU prefers hostnames. It does not play well with IP address instead of hostname.

Please change the policy to client hostname and try again.

[bluedemon]

Unfortunately we having 2 server with same hostname. One is the Unix server (192.168.10.254) while another one is a Windows server (192.168.10.202). Currently i cannot rename the hostname for both server as it is related to certain software.

If I use hostname in the policy, it detect the Windows server instead of the Unix server. I already added an entry in the Master server host file to point seemglv to 192.168.10.254 but the netbackup will keep on detect the Windows server

[Marianne]

This seems to be the cause of the problem.

NBU relies on forward and reverse name lookup.

The NBU agent on the client is sending the backup request with the local hostname and IP address to the master.
If the master is on DNS (Windows masters normally are) then it will use DNS for hostname resolution.

Fix the duplicate hostname issue. 
I cannot see how anything is going to work while there are duplicate hostnames in the environment.

[sdo]

There might be a workaround, as long as both clients do not have database agent backups.

The Unix host requires Oracle RMAN agent backups.  Does the Windows host require any DB agent style backups?

[RiaanBadenhorst · Accepted solution]

You can try remove the entry from your master hosts file. Then on the client, change the CLIENT name in bp.conf to the IP address. That should ensure everything is happening via IP.

Finally, you might also try and disable reverse loopup on the master server properties. Be aware that changing this might affect other working clients.

[bluedemon]

I only use the Netbackup to backup the Oracle Server. We do not need to backup the Windows server. In fact, the Windows Server does not have any Netbackup agent in it