Error ORA-1114 on Heterogeneos System Copy

[Former Member]

Hi everybody!

Someone can help me? I got the next error when we are trying to make an Heterogeneos System Copy. I can't continue the export from the installation master. The next is the error:

Compiled Jan 14 2006 06:36:14

/sapmnt/BWP/exe/R3load -datacodepage 1100 -e /tmp/sapinst_instdir/NW04/COPY/EXPORT/ABAP/COPY/NUC/DBEXP/SAPDODS.cmd -l /tmp/sapinst_instdir/NW04/COPY/EXPORT/A

BAP/COPY/NUC/DBEXP/SAPDODS.log -stop_on_error

(DB) INFO: connected to DB

(DB) INFO: DbSlControl(DBSL_CMD_NLS_CHARACTERSET_GET): WE8DEC

Syslog: k CPU : UMGSETTING&TRANS&UNICONV& &'N','U'&Warning& rscpexcc 29

Syslog: k CPU : UMGSETTING&EINDX&UNICONV& &'N','M'&Warning& rscpexcc 25

(rscpMC) Warn: UMGSETTINGS says: RADCUCNT not succesful

(GSI) INFO: dbname = "BWP20060217060223 "

(GSI) INFO: vname = "ORACLE "

(GSI) INFO: hostname = "srm-pro "

(GSI) INFO: sysname = "SunOS"

(GSI) INFO: nodename = "srm-pro"

(GSI) INFO: release = "5.9"

(GSI) INFO: version = "Generic_122300-15"

(GSI) INFO: machine = "sun4u"

(GSI) INFO: instno = "0020221516"

DbSl Trace: Error 1114 in stmt_fetch()

DbSl Trace: ORA-1114 occurred when executing SQL statement (parse error offset=0)

(EXP) ERROR: DbSlExeRead failed

rc = 99, table "/BIC/B0000022000"

(SQL error 1114)

error message returned by DbSl:

ORA-01114: IO error writing block to file %s (block # %s)

(DB) INFO: disconnected from DB

/sapmnt/BWP/exe/R3load: job finished with 1 error(s)

/sapmnt/BWP/exe/R3load: END OF LOG: 20080209205651

Thanks

Best Regards

[Former Member]

Hi Sergo!

The OS is a Solaris 9 and the directories are you talking about have the privileges that you say. Also we searched in the Oracle Metalink and we found that the error coud be for a full space in the PSAPTEMP tablespace. Our database is about 400GB and we get in our PSAPTEMP 17GB free. Do we need more space?

Thanks

Best regards

[markus_doehr2]

depending on the number of parallel loads running with each using PSAPTEMP it may be advised to increase the size (and later reduce it again).

Honestly, to me this appears more a problem of the underlying I/O subsystem. What filesystem do you use? What's the value of the parameter "filesystemio_options"?

Apart from that I would suggest you use the LATEST version of R3load (yours is 2 years old).

Markus

[Former Member]

Hi Markus

The O.S team toll me that they use the defualt options to create FS(ufs) and for the I/O too.

As well we will are updated the R3**** and will try again.

Best Regards

Reynaldo Rebolledo

[markus_doehr2]

Check the parameter is "setall" (will give maximum throughput on UFS).

Markus

[Former Member]

HI Reynaldo Rebolledo Romero . When we made a upgrade our base was 190 GB, the PSAPTEMP is 15 Gb,

The turned out files 12.5 GB ,It is desirable for you to increase PSAPTEMP at least more than 20 GB and if that is possible still.

Thanks and regards. Reward helpful answer.

Edited by: Sergo Beradze on Feb 13, 2008 6:33 AM

[former_member204746]

it looks liek you have a corrupted Oracle block:

ORA-01114: IO error writing block to file %s (block # %s)

ask your OS folks to check if you have disk problems.

[Former Member]

Hi everybody!

We have solved the problem. We need to rebuild the oracle temporary tablespace and resize it about 100GB.

Thanks to all for your help.

Best regards.

[Former Member]

I'm dont know SUNOS , in HP-UX :

Check whether the "oracle" file exists on $ORACLE_HOME/bin, has a size other than 0 and has the "-rwsr-s--x" authorizations. The owner must be ora<sid>. The group is dba.

If the file is empty, you can create it again by relinking the software in accordance with Note 97953.

You can assign authorizations and the owner/group as follows:

chmod 6751 oracle

chown ora<sid> oracle

chgrp dba oracle

Also check the authorization of the $ORACLE_HOME/bin directory. This should be set to "drwxr-xr-x".