Oracle RAC RMAN Backup

[Patrick_129]

Hi guys,

I have two servers running on two servers.
physical instance - P1 and P2
Virtual instance - V1

TNS was configured as in such a way that when you execute RMAN or sqlplus to the physical instance, it will connect to the virtual instance automatically.

My problem is when i try to manually run RMAN on either server, sometimes i get the following error...
-------------------------------------------
RMAN-00601: fatal error in recovery manager
RMAN-03004: fatal error during execution of command
RMAN-10038: database session for channel default terminated unexpectedly
-------------------------------------------

My recovery catalog server is a linux server.

Below is my command i used to connect the database and catalog...

--------------------------------------------
RMAN> connect catalog rman/XXXXX@BKUP
connected to recovery catalog database
RMAN> connect target sys/XXXXXX@V1
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-00601: fatal error in recovery manager
RMAN-03004: fatal error during execution of command
RMAN-10038: database session for channel default terminated unexpectedly
--------------------------------------------

If i tried connect the database first before the catalog or vice-versa, the result is the same.

The strage thing is that it only happened occasionally.

Any idea???
Thans for the help

patrick

[Henk Geurts]

the error codes don't explain much:
RMAN-00601: fatal error in recovery manager
Cause: A fatal error has occurred.
Action: This message should be accompanied by other error message(s) indicating the cause of the error.
RMAN-03004 fatal error during execution of command
Action: This message should be accompanied by one or more other error messages indicating the cause of the error.
Action: Check the accompanying errors.
RMAN-10038: database session for channel num terminated unexpectedly
Cause: The database connection for the specified channel no longer exists. Either the session was terminated by some external means or the channel terminated because of an internal error.

seems to be a connection problem somewhere. any message in the syslog?

regards.

[Indira Aramandla · Accepted solution]

Hi Patrick,

Errors like RMAN backup fails with errors with RMAN-00601: fatal error in recovery manager, RMAN-3004: fatal error during execution of command and RMAN-10038: database session for channel %s terminated unexpectedly
The cause could be the database initialization parameter SHARED_POOL_SIZE value was defined too low.
The fix would be to increase the database initialization parameter SHARED_POOL_SIZE.

The other area to check for when encountering errors like RMAN-601 RMAN-3004 is:-

you must set a mandatory parameter in the target database's "init.ora" in order
for RMAN to successfully backup the database. Set DB_DOMAIN in your "init.ora" to your Net8 domain name. DB_DOMAIN should match the NAMES.DEFAULT_DOMAIN parameter in your "sqlnet.ora" file.

Example:

db_domain = world

You must set DB_DOMAIN in the target database init.ora file.

Indira A

[Patrick_129]

Hi,

There no error in syslog or any Oracle trace file as informed by my UX and DB team.

I belived it something got to do with the TNS because if i do a RMAN to the physical instance (P1 or P2) and then connect to the catalog, it's OK! :) it is only when i connect to the virtual instance and the catalog, i get the error message but again, i happened occasinally.

Question,
I'm using netbackup in this case, should i connect to the pysical instance in the RMAN script to backup the virutal instance(V1)?

I believed it the same thing because connect using P1 , the TNS will connect to V1 automatically.

Anyway, it using OCFS on a shared storage.

Thanks alot ! :)

[Patrick_129]

Thanks indira.

I get my Oracle team to take a look at the
db_domain parameter...

[Henk Geurts]

Patrick
please give some points to Indira as he really helped you .......

assigning points is the way to go....

regards.

[Patrick_129]

Hi Henk,

Thanks for informing, Henk.
I need to get my Oracle team verify first before closing the thread. I know the protocol here ;)

[Duncan Edmonstone]

Patrick,

The following is from the RAC Admin guide:

Connection to the Target Database in a Real Application Clusters Configuration
When you start RMAN and connect to the target database, RMAN can only connect
to one instance in a Real Application Clusters database at a time. Note that this
connection is a utility connection that does not perform any backups or restores,
and applies only to the connection made from the RMAN command line.
Assume that node1, node2, and node3 are net service names for three instances in a
Real Application Clusters configuration. In this case, connect to the target database
with only one of these net service names. For example, you can connect as follows:
% rman TARGET SYS/oracle@node2 CATALOG rman/cat@catdb

So I expect by connecting to the virtual address you are breaking this rule.

HTH

Duncan

[Brian Crabtree]

Patrick,

You should also configure your backup to connect to each node to backup archivelogs as well.

Brian

[Patrick_129]

I had changed the script to backup the physical instance rather than the virtual instance.

Thanks for all input. I will know the result tommorrow :)

patrick