Oracle backup failures..

[RamNagalla]

have a two Oracle DBs on the solaris local Zone.

both  are iniciated backups from the Rman script to take the backups or Oracle DB and archive logs..

when Rman Fires the backups one DB is getting successfull and another DB is getting failues with EC 54.. no matter what time they are starting..

NB version in master and media and client are same its 6.5.6

the local zone has the OS solaris 8

I dont think its a issue with the network.. as one DB is running successfull

out of 2 DBs only one is failling.

DBclient log:-

00:18:28.303 [14628] <8> connectSock: WRN - nb_connect_to_vnetd failed for <mediaservername>: 25
00:18:43.297 [14628] <16> readCommFile: ERR - timed out after 5 seconds while re
ading from /usr/openv/netbackup/logs/user_ops/dbext/logs/14628.0.1305517467
00:18:43.297 [14628] <16> serverResponse: ERR - cannot connect on DATA socket <mediaservername>: .IPC:/tmp/vnet-21054305523099879243000000000-NJaWhP
;2dc8898194891be7b7c975cd5e3b657b;4;1800
00:18:43.297 [14628] <16> CreateNewImage: ERR - serverResponse() failed
00:18:43.297 [14628] <16> VxBSACreateObject: ERR - Could not create new image wi
th file /al_s29851_p1_t751243466.
00:18:43.297 [14628] <16> xbsa_CreateObject: ERR - VxBSACreateObject: Failed wit
h error:
   Server Status:  Communication with the server has not been iniatated or the s
erver status has not been retrieved from the server.
 

Please let me know how can I proceed with it..

[Yogesh9881]

when Rman Fires the backups one DB is getting successfull and another DB is getting failues with EC 54

As per the above line i understand backup is not happening with RMAN for another DB.

But as per the error logs .....

<16> readCommFile: ERR - timed out after 5 seconds while re

You need to increase default time out value from master / media / client host properties.

[RamNagalla]

for both DBs. Rman is firing the backups.....

and both are reaching to the master server( i could see those in activity moniter) one  DB is getting sucessfull.. and other is failing..

i have tried with the time out vaules..with no luck..

still i am not getting any idea.. how one DB can fail when one other DB with same setting running fine..

i am suspecting something wrong from the DB or  Rman side.. but  i am not strong to dig in Rman or DB side..

any ideas...? so that I can go ahead further.....

[Yogesh9881]

if you are not sure about Script then

You can logon to Client then run /usr/openv/netbackup/bin#./jbpSA

(you need X services on client).

then select DB that you want to backup and create script (easy wizard for creating script)

locate newly created script in policy start backup from master server.

[watsons]

What Yogesh suggested is the correct direction.

How much data are you backing up, for the successful job and the failed job?

If the failed job has more data, it's very likely a timeout problem. From logs it timeout right after 15min.

[RamNagalla]

i need to access the client remotly.. where  i  can not use the jnbSA...

i only can get the command line interface....

[RamNagalla]

anyone have any suggestions  to dig this issue..

I am looking forward for the path to troubleshoot..

I have tried with timeout valules....

[Marianne]

"connectSock: WRN - nb_connect_to_vnetd failed for <mediaservername>: 25"

Is you master and media server the same server?

At this point, the client is actually trying to connect to bprd (via vnetd) on the master server.

Verify that the 1st SERVER entry in the client's bp.conf is the master server.
Verify correct hostname entry for Master server in client's hosts file.

Also ensure /etc/hosts entry on master server for this client.

Check master's bprd log file (create the folder and restart NBU if it does not exist) for incoming request from client's IP address. Check the log for master's attempt to resolve IP address to client hostname that corresponds with client hostname in the Oracle policy.

[RamNagalla]

Master and Media servers are different..

first entry in the client bp.conf is master server bp.conf only..

we are using the DNS servers.. so no /etc/hosts issue..

I have verrified the servers entries in nslookup and bpclntcmd -hn ,-ip in both master and client they are giving the same result..

and in the bprd log.. the same ip address is trying to communicate. that is the host name configured in policy..

I would like to stress on thing hear.. the same client box having 2 Oracle DB policys.. one policy is getting successfull.. but other is failing with 54.

anything that I am missing or I can check..

[Marianne]

Have you tried to increase Timeouts on master and media server?

For database backups, we start with Connect and Read Timeouts of 1800.

[emret]

after that you can modify the oracle script like this ;

CMD_STR="

ORACLE_HOME=$ORACLE_HOME

export ORACLE_HOME

ORACLE_SID=$ORACLE_SID

export ORACLE_SID

$RMAN target $TARGET_CONNECT_STR nocatalog msglog $RMAN_LOG_FILE append << EOF

RUN {

ALLOCATE CHANNEL ch00 TYPE 'SBT_TAPE';

ALLOCATE CHANNEL ch01 TYPE 'SBT_TAPE';

SEND 'NB_ORA_POLICY=Daily_Backup_xxxxxxx, NB_ORA_SERV=xxxxxxxx, NB_ORA_CLIENT=xxxxxxxx’;

BACKUP

$BACKUP_TYPE

SKIP INACCESSIBLE

TAG hot_db_bk_level0

FILESPERSET 5

# recommended format

FORMAT 'bk_%s_%p_%t'

DATABASE;

sql 'alter system archive log current';

RELEASE CHANNEL ch00;

RELEASE CHANNEL ch01;

# backup all archive logs

ALLOCATE CHANNEL ch00 TYPE 'SBT_TAPE';

ALLOCATE CHANNEL ch01 TYPE 'SBT_TAPE';

BACKUP

filesperset 20

FORMAT 'al_%s_%p_%t'

ARCHIVELOG ALL DELETE INPUT;

RELEASE CHANNEL ch00;

RELEASE CHANNEL ch01;

[RamNagalla]

I have kept the connect and read timeout as max as I can possible.. its around 32000..

[RamNagalla]

i am not clear with the tnsnames.  I will have to check with my DBA for this..

[emret]

.

[RamNagalla · Accepted solution]

finally my case endup with symantec call....

they find that its a issue with the linking of Rman to the netbackup agent...once we corrected that it starts working fine...

thank you very  much for alll for your suggestions...:-)