cancel
Showing results for 
Search instead for 
Did you mean: 

URGENT: DBIF_REPO_SQL_ERROR - not even OSS has been able to help!

former_member193337
Participant
0 Kudos

Hello Experts!

Since May 1th we have been having a very serious problem with the general performance generating dump code ORA-03113 and ORA-3114 which has grown constantly. We have had to reinitiate the system repeatedly.

The following activities were made:

 Replacement of a damaged disc of a volume used by the server located

in SAN, which generated a process of reconstruction of RAID 5

 Approximately 2 Days after the process had been initiated there's a problem of slowness and disconnection of users generating the following errors:

- DBIF_NTAB_SQL_ERROR

- DBIF_RTAB_SQL_ERROR

- DBIF_REPO_SQL_ERROR

With the following detail:



ORACLE error 3113 encountered ORA-03113: communication end-of-cases

out on channel.

ORACLE error 3114 encountered ORA-03114: not connected to ORACLE.

 We have increased the Redo log to 100 Mb. Alse we have reorganized table CE4VE01. Possibility of corrupt blocks in the data base was verified.

 We have Installed local additional discs to the server for the handling

of the paging files that they were in SAN

 The network cables were changed, they were changed to different ports, and extended to 100 Mbps Full Duplex in switch

 We've changed the Network cards by new ones of the

same original model

 The Drivers were updated and we verified the presence of the latest firmware

 We have made tests of stress to all hardware, which indicated

that there were no problems in the components

At the moment the infrastructure of communications was OK, but the errors continued as well as the shutdowns of the system

We have also followed the OSS instructions described below:

We setted all the database parameter as per the SAP note 124361.

We applied the latest oracle patch level as per the SAP note 539921.

Thanks in advace for your help!,

Frinee

Accepted Solutions (0)

Answers (1)

Answers (1)

markus_doehr2
Active Contributor
0 Kudos

> ORACLE error 3113 encountered ORA-03113:

> communication end-of-cases

> out on channel.

> ORACLE error 3114 encountered ORA-03114: not

> connected to ORACLE.

Those errors indicate, that the shadow process was terminated.

What do you see in your Oracle alert<SID>.log when this error occurs? What's in the corresponding .trc file in saptrace/usertrace?

--

Markus

former_member193337
Participant
0 Kudos

Hello Markus, Thanks for your reply!

The Oracle alertPRD.log doesn't show any error and the .trc file is empty it does not generate any trace...

We are very concern as our production environment it's been down all morning and the operation is stopped! :'(

Thanks!

Frinee

markus_doehr2
Active Contributor
0 Kudos

If Oracle doesn't write any alerts and the tracefiles are empty... is the filesystem full? I have never seen any ORA database, that doesn't write any tracefiles...

I'd escalate the message (raise the prio to very-high), if that does not help, I'd call the local customer support every 30 minutes about a status update...

--

Markus

former_member193337
Participant
0 Kudos

Thanks Markus, we will do that!!!, this is very strange...