Datapump Network Import from Oracle 10.2.0 to 10.1.0

I have self-developed table bulk copy system based on dmbs_datapump API. It is pretty stable and worked very well for a long time on many servers with oracle 10.1.0. But some days ago we have installed new oracle server with oracle 10.2.0 and have noticed, that our system can't copy tables from new server to another servers. After some time of debugging we got description for our situation... dbms_datapump.open() call is failing with "ORA-39006: internal error" exception. dmbs_dapatump.get_status() returns more detailed description: "ORA-39022: Database version 10.2.0.1.0 is not supported."... We're really disappointed... what we can do to use dbms_datapump.open on 10.1.0 to copy table from 10.2.0?
P.S. We have tried to use different values (COMPATIBLE, LATEST, 10.0.0, 10.1.0) for version parameter of open()... but without luck...
Please, help, because we can't upgrade such many old servers to 10.2.0 and we really need to use this new server...
Thanks in advance,
Alexey Kovyrin

Hello,
Your problem is about the Character Set.
WE8ISO8859P1 stores character in *1* byte but AL32UTF8 (Unicode) uses up to *4* bytes.
So, when you Import from WE8ISO8859P1 to AL32UTF8 you may have the offending error ORA-12899.
To solve it, you may change the option of the Datatype of your column. And choose the option CHAR
instead of BYTE (the default), for instance:
VARCHAR2 (100) --> VARCHAR2(100 CHAR)Then, you could import the data.
You may also change the parameter NLS_LENGTH_SEMANTICS to CHAR.
alter system set nls_length_semantics=char scope=spfile;Then, restart your database.
NB: Although the parameter nls_length_semantics is dynamic you have to restart the database so
that it's taken into account.
Then, you create the Tables (empty) and afterwards you run your Import.
Hope this help.
Best regards,
Jean-Valentin
Edited by: Lubiez Jean-Valentin on Mar 21, 2010 4:26 PM

Similar Messages

Datapump network import from 10g to 11g database is not progressing

We have an 11.2.0.3 shell database in RedHat Linux 5 64-bit platform, and we are pulling data from its 10g database (10.2.0.4) HP-UX platform using datapump via NETWORK_LINK. However, the import seems to be not progressing and completing. We have left it run for almost 13 hours and did not even import a single row to the 11g database. We even try to import only one table with 0 rows but still it is not completing, the logfile just continue to be lopping on the following:
Worker 1 Status:
Process Name: DW00
State: EXECUTING
Estimate in progress using BLOCKS method...
Job: IMP_NONLONG4_5
Operation: IMPORT
Mode: TABLE
State: EXECUTING
Bytes Processed: 0
Current Parallelism: 1
Job Error Count: 0
Worker 1 Status:
Process Name: DW00
State: EXECUTING
We also see this:
EVENT SECONDS_IN_WAIT STATUS STATE ACTION
SQL*Net message from dblink 4408 ACTIVE WAITING IMP_NONLONG4_5
Below is our par file:
NETWORK_LINK=DATABASE_10G
DIRECTORY=MOS_UPGRADE_DUMPLOC1
LOGFILE=imp_nonlong_grp-4_5.log
STATUS=300
CONTENT=DATA_ONLY
JOB_NAME=IMP_NONLONG4_5
TABLES=SYSADM.TEST_TBL
Any ideas? Thanks.

Thanks a lot for all who have looked and responded to this, I appreciate you giving time for suggestions. As a recap, datapump export and import via dumpfile works on both 10g and 11g databases, we are only having issue pulling data from the 10g database by executing datapump network import (using dblink) from the 11g database.
SOLUTION: The culprit was the parameter optimizer_features_enable='8.1.7' that is currently set on the 10g database, we have taken that out of the parameter file of the 10g database and network import worked flawlessly.
HOW DID WE FIGURE OUT: We turned on a trace to the datapump sessions and trace file shows something about optimizer (see below). So we have taken one by one any parameter that has something to do with optimizer and found out it was optimizer_features_enable='8.1.7' parameter that when removed makes the network import successful.
SELECT /* OPT_DYN_SAMP */ /*+ ALL_ROWS opt_param('parallel_execution_enabled',
'false') NO_PARALLEL(SAMPLESUB) NO_PARALLEL_INDEX(SAMPLESUB) NO_SQL_TUNE
*/ NVL(SUM(C1),0), NVL(SUM(C2),0), NVL(SUM(C3),0)
FROM
(SELECT /*+ NO_PARALLEL("SYS_IMPORT_TABLE_02") INDEX("SYS_IMPORT_TABLE_02"
SYS_C00638241) NO_PARALLEL_INDEX("SYS_IMPORT_TABLE_02") */ 1 AS C1, 1 AS C2,
1 AS C3 FROM "DBSCHEDUSER"."SYS_IMPORT_TABLE_02" "SYS_IMPORT_TABLE_02"
WHERE "SYS_IMPORT_TABLE_02"."PROCESS_ORDER"=:B1 AND ROWNUM <= 2500)
SAMPLESUB
SELECT /* OPT_DYN_SAMP */ /*+ ALL_ROWS IGNORE_WHERE_CLAUSE
NO_PARALLEL(SAMPLESUB) opt_param('parallel_execution_enabled', 'false')
NO_PARALLEL_INDEX(SAMPLESUB) NO_SQL_TUNE */ NVL(SUM(C1),0), NVL(SUM(C2),0)
FROM
(SELECT /*+ IGNORE_WHERE_CLAUSE NO_PARALLEL("SYS_IMPORT_TABLE_02")
FULL("SYS_IMPORT_TABLE_02") NO_PARALLEL_INDEX("SYS_IMPORT_TABLE_02") */ 1
AS C1, CASE WHEN "SYS_IMPORT_TABLE_02"."PROCESS_ORDER"=:B1 THEN 1 ELSE 0
END AS C2 FROM "DBSCHEDUSER"."SYS_IMPORT_TABLE_02" "SYS_IMPORT_TABLE_02")
SAMPLESUB

Importing from Oracle Designer

Hi, I am looking to migrate from Oracle Designer. We have 2k plus objects in a enterprise model supporting multiple applications. When following the dialogs to import into DM 3.1, I connect to the repository fine but when I select our Logical application system, I am not seeing all our entities. What is the basis for selecting entities to present? Do they need to be checked in? We have some that are not yet versioned that do not show up and I can see them not being there but others that are versioned do not show up. I think most of these are checked out but I see some that are checked out as well. Any ideas, thks.
Edited by: 930485 on Apr 26, 2012 7:50 AM

Yes they need to be checked in if they are versioned.

Import from Oracle 8.1.6-7 to 8.1.5

Hi folks
I'm having this problem: a software supplyer from our company converted to us their proprietary database used inside a HR management system to an Oracle DB. They send to us the dmp file, but when we try to import we've got this:
IMP-00010: not a valid export file, header failed verification
IMP-00021: operating system error - error code (dec 2, hex 0x2)
I checked at the raw dump file it was generated with EXPORT:V08.01.06F. In our company we use Oracle 8.1.5 whose export produces at top EXPORT:V08.01.05. Does anybody know if is it possible to download only import from above version and make it work? Suggestions are welcome...
null

We have the same problem ?
Is there a solution in the meantime ?
TIA!
Michael

Error while table Import from Oracle (OBIEE server 64bit & Client 32bit)

Hi All,
Problem : ODBC/OCI Connectivity Issue between OBIEE 11.1.1.6.2 BP1 (64-bit) server and OBIEE 11.1.1.6.2 BP1 Client (32-bit) local machine.
We have an OBIEE 11.1.1.6.2 BP1(64-bit) setup on a 64-bit ‘Windows Server 2008 R2 Enterprise’ machine.
OBIEE dev team members have 32-bit local machines.
OBIEE “Admin team member” (who will work on RPD) has OBIEE 11g client (BP1) installed on their *32-bit local machine*(which is common client software for 32/64 bit.)
Now, two ODBC connections were created on this 32-bit local Client machine, with OBIEE 11g client installed on it.
One with “BI server driver” for connection to OBIEE server machine(64-bit) to open RPD and other with “Oracle database driver” to connect to the Oracle database on 64-bit machine, to import tables.
With this setup done, I am able to open the RPD in online mode and view the existing sampleapp related objects present in it.
But when I try to import metadata from OBIEE client admin tool using OCI or ODBC, it gives an error that “Connection fails! “
I have the required tns entry in my tnsnames.ora file and I am able to connect to this database from the local client machine, using sqlplus.
Kindly let me know if there are any restrictions in using this type of a setup combination of 32-bit and 64-bit machines for OBIEE 11.1.1.6.2 BP1 setup.
If not then, what steps are to be followed to do the setup btw 64-bit OBIEE server and 32-bit OBIEE client to make table import working.

Hi Veeravalli.. Thanks for your response.
I have my tns file residing in both the location you mentioned on the server machine. (64-bit)
•     {Oracle_BI1}\network\admin directory
•     {oracle_common}\network\admin directory
And also in the below Oracle_home path of oracle_client in the local machine. (32-bit, Oracle 11g R2 database client)
•     {oracle_Home}\product\11.2.0\client_1\network\admin
Also, I have used the full expanded tnsnames as you mentioned. Still issue persists.
When I use , “DataDirect 6.0 ODBC Oracle Wire Protocol” driver, for creating the system DSN and try to import tables from OBIEE client admin tool on local client machine, the connection happens and table list comes up, but it’s an SAP product and hence it throws a msg to purchase the tool to go further with table import.
I am wondering if I any other configuration steps need to be carried out, with the ODBC/OCI set up between 64-bit OBIEE server and 32-bit OBIEE client.
Also, the source database that I am trying to connect to is yet again on a 64-bit machine(Oracle 11g R2).
Please suggest.

How to import from Oracle to ODI View query definition?

The scenario:
I have a View in Oracle database with name 'XV'.
When I loaded metadata from this Oracle database the 'XV' structure was imported too.
Is it possible to get in ODI the 'XV' query definition in format 'CREATE VIEW ... ' or 'SELECT * FROM'?
Thanks a lot for answers.

Still cannot connect.
The Administration Tool help suggests the tnsnames.ora is not required:
"For Oracle Database data sources, the data source name is either a *full connect string* or a net service name from the tnsnames.ora file. If you enter a net service name, you must ensure that you have set up a tnsnames.ora file within the Oracle Business Intelligence environment, in ORACLE_HOME/network/admin".
I've tried the tnsnames.ora route, but this still does not work. I created a tnsnames.ora file in <MW_HOME>\Oracle_BI1\network\admin. Is this correct?
I'm assuming that I don't need an Oracle client install on the same (Windows) PC that I've got Oracle BI 11g installed on. Is this correct?

Oracle 11g- Straing behaviour of query after importing from Oracle 10g

Hi,
I have a table in Oracle 10g as follows:
Create Table xyz (col1 varchar2(50), col2 varchar2(50));
With following Data
Col1     Col2
A     320
A     110
A     290
A     380
B     ABC
B     256
B     LMN
I am running following Query
select * from xyz
Where Col1='A' and Col2=110
It works fine. But when I export this table and import it in Oracle 11g. It says invlid identifier.
But if I enclose 110 in single quotes. It works fine.
Also If I recreate this table in Oracle 11g like
Create table xyz1
as select * from xyz;
Now alos I am able to run this query smoothly.
select * from xyz1
Where Col1='A' and Col2=110
What is wrong exporting this table from 10g to 11g.
Any comments/suggestion??
Aarbi

The check in your where clause
Col2=110Is comparing a string (Col2 is defined as a VARCHAR) with a numeric literal, so there will be an implicit conversion taking place from character to number. The query then fails due to the B LMN row when 'LMN' fails number conversion.
I'm guessing there was there an index on the table in your 10g installation which would allow the query to be satisfied without checking the B ABC or B LMN rows but is not present or not used in the 11g installation so a full table scan results in an attempt to convert 'ABC' and 'LMN' to a number. Check the explain plans.
Or it could even just be a difference in the order in which the two conditions in the where clause are evaulated between the two versions.
The solution, as you have already found is to do a string comparision
Col2='110'Edited by: Cyn on Dec 7, 2009 12:38 PM

Data Import from Oracle Dump to SQL Server(2000/2005)

Hi Friends,
I'm using Oracle 10g.
I've a scenario.
I've Oracle Dump file. Now i would like import in SQL Server(2000/2005).
Can any one suggest me how to do this?
Any idea,clue will be highly appreciated.

I think you will need to import the data into SQL Server from an ASCII flat file using SQL Server DTS if avialable. You can use SQL Plus to extract the data outof Oracle into comma delimited flat files and then use SQL Server DTS load into the SQL Server database.

SQL Developer Data Modeling - import from Oracle Designer Model

Hi,
I do want to try to import model from a Designer repository.
The first step is to create connection to the repository (9i version).
I have created the one, however any attempt to test the connection or to go to the next step simply finishes with error "4", without any valueable message.
Any idea or suggestion ?
Thanx.

A clarification of the Data Modeling feedback application and this forum. The developers are part of the SQL Developer development team, but as for all our features, each developer has a focus area, so the Data Modeling developers will tend to answer those questions. There will not be a separate Data Modeling forum, so once we're production all SQL Developer Data Modeling questions will be on this forum. We want to encourage all users to get into the habit of tagging posts. Having the posts tagged means that users can skip them if they're not their focus area and users can search using the tags. The more tags are used in all the forums the better. You may well find find the answers to your questions are outside this forum!
As SQL Developer Data Modeling in in its first early adopter phase, we have provide a feedback application, where users can provide feedback, log bugs and ask questions. We did the same for SQL Developer before it's initial production release and found this to be very useful for the product.
I'm happy for "How To" questions to be on this forum, as the answers benefit the broader community.
Sue Harper

Import from Designer crashes with IMAGE data type

Hello,
I'd like to report an issue with Data Modeler while importing from Oracle Designer. If you consider this a bug, please let me know if I need to do anything else to request a fix for it.
I've used the import > Oracle Designer Model tool to import an application model. The import runs fine for a while, then crashes with a message "There are errors in import - check log file".
The application to be imported has 509 domains, 963 entities to import.
The summary info displayed after the import (when I close the error message pop up) is:
All statements: 1472
Imported statements: 538
Failed statements: 0
Not recognized statements: 934
The log file has the lines below:
2011-04-28 10:55:46,505 [Thread-8] ERROR XMLTransformationManager - Unable to load object from XML: C:\Users\Beatriz\Documents\ODTUG\2011\SDDM designs\GAME\GAME\rel\0EEBE15E-D9EFB22E4D6B\subviews\02A206F8-0C6C-0DF9-E643-5919DBCB895A.xml
java.io.FileNotFoundException: C:\Users\Beatriz\Documents\ODTUG\2011\SDDM designs\GAME\GAME\rel\0EEBE15E-D9EFB22E4D6B\subviews\02A206F8-0C6C-0DF9-E643-5919DBCB895A.xml (The system cannot find the file specified)
     at java.io.FileInputStream.open(Native Method)
     at java.io.FileInputStream.<init>(FileInputStream.java:106)
     at oracle.dbtools.crest.model.metadata.XMLToObjectTransformer.transformToObject(Unknown Source)
     at oracle.dbtools.crest.model.metadata.XMLTransformationManager.transformFromXMLToObject(Unknown Source)
     at oracle.dbtools.crest.model.metadata.XMLTransformationManager.openDesignPart(Unknown Source)
     at oracle.dbtools.crest.model.design.Design.openDesign(Unknown Source)
     at oracle.dbtools.crest.swingui.ControllerApplication$Recent$1.run(Unknown Source)
2011-04-28 11:03:44,114 [Thread-12] WARN ODODomain - ODOType.initStructuredTypeAttributes() - Unable to find Logical DataType for: IMAGE
2011-04-28 11:03:46,532 [Thread-12] ERROR ODExtractionHandler - Error during import from Designer Repository
java.sql.SQLException: Numeric Overflow
     at oracle.jdbc.driver.NumberCommonAccessor.throwOverflow(NumberCommonAccessor.java:4380)
     at oracle.jdbc.driver.NumberCommonAccessor.getInt(NumberCommonAccessor.java:111)
     at oracle.jdbc.driver.OracleResultSetImpl.getInt(OracleResultSetImpl.java:928)
     at oracle.jdbc.driver.OracleResultSet.getInt(OracleResultSet.java:434)
     at oracle.dbtools.crest.imports.oracledesigner.logical.ODOEntity.generate(Unknown Source)
     at oracle.dbtools.crest.imports.oracledesigner.ODExtractionHandler.generateDesign(Unknown Source)
     at oracle.dbtools.crest.imports.oracledesigner.ODExtractionController$Runner.run(Unknown Source)
     at java.lang.Thread.run(Thread.java:619)
I see the issue with the IMAGE data type in the log. The model to be imported does have domains and entities using this data type. The domain actually did get imported, but it shows as "Unknown" logical type, even though there is an Image type available in the Data Modeler type list. None of the entities using IMAGE got imported.
Please let me know if you can provide a fix or workaround for this (other than changing the format in Designer).
Thank you,
Beatriz.

Hello Beatriz,
definitely this is a bug. The problem with entities is not IMAGE data type it's supported there (it's overlooked for domains). The problem is volume information - Initial, Maximum, Average, Growth rate. No workaround for that.
Domain can be changed manually if it's just one affected.
Philip

Importing Statistics - Oracle 11g upgrade

Hi,
We are in the middle for planning for migration of oracle 9.2.0.8 database hosted in HP-UX to oracle 11gR2 to Linix. The database size is 1 TB ( 400GB of table data & 600GB of index data).
Please let us know whether we can use the option of import/export the statistics from the Oraclr 9i to oradle 11g database. The database is highly OLTP/Batch. Will it be any Query performance problem due to the statistics import from oracle 9i to oracle 11g.
Any suggestion are welcome and let me know if you need any more information.
thanks,
Mahesh

Hello,
Please let us know whether we can use the option of import/export the statistics from the Oraclr 9i to oradle 11g databaseIf I were you, when the data are imported to the 11g Database I would refresh the Statistics by using the dbms_stats Package.
Then, you can test the most common queries on the new Database. If some perfomance troubles appear, you can use classical tools (Statspack, Explain Plan, TKPROF,...) or some useful tools you have in 11g as SQL Tuning Advisor or SQL Access Advisor :
http://download.oracle.com/docs/cd/E11882_01/server.112/e16638/sql_tune.htm#PFGRF028
http://download.oracle.com/docs/cd/E11882_01/server.112/e16638/advisor.htm#PFGRF008
Hope this help.
Best regards,
Jean-Valentin

Upgrade from Oracle 8i in Win98 to Oracle 11g in Window Vista

Hi,
I have an existing system that runs in Window98 using Oracle 8i as database. We are planning to convert the system from using Oracle 8i to 11g, and this new converted system will be running in Window Vista.
What I have in mind is doing export and import from Oracle 8i to 11g, but it is quite troublesome. Is there anyway I can upgrade the db to 11g in Window Vista?

Hi,
When upgrading from 8i to 11g, you cannot perform direct upgrade anyway.
If you want to upgrade (not using exp/imp), you can install 9i (9.2.0.4 and up) on the win98 machine, and upgrade the database to 9i.
Then you can take the database files, copy them (like cold backup) to the vista server, and you will able to upgrade it to 11g.
Note that you cannot use the DBUA in this scenario, you will have to upgrade it manually.
Manual upgrade: [http://download.oracle.com/docs/cd/B28359_01/server.111/b28300/upgrade.htm#i1011600]
Liron Amitzi
Senior DBA consultant
[www.dbsnaps.com]
[www.orbiumsoftware.com]

4.0 EA3 import from designer and database leads to differences in compare models

Hello,
I followed two ways to import a model in Data Modeler:
1 - an import from Oracle Designer
2 - an import from the database generated on basis from Oracle Designer
Then I compare the two models in Data Modeler, and found differences ...
One of the differences is in Table check constraints. The text is the same when viewing it in the compare model.
So I opened the corresponding XML files for the tables and compared the XML for the constraint text. And there are differences:
example:
<rule>(file_locatie is not null and file_naam_org is not null) or<br/>(file_locatie is null and file_naam_org is null)</rule>
<rule>(file_locatie is not null and file_naam_org is not null) or<br>(file_locatie is null and file_naam_org is null)</rule> <<-- designer
the design imported from designer contains br, the databse imported version contains br/
This should not be considered during comparision, or better: boths imports should use the same way of formatting.
Joop

The Rule specification is not re-formatted in Data Modeler code; it is stored as extracted from the database.
For Oracle Database, the Rule is extracted using SQL of the form SELECT search_condition FROM dba_constraints/all_constraints WHERE constraint_type = 'C';
For Oracle Designer, the Rule is extracted using SQL like SELECT txt_text FROM rm_text_lines WHERE txt_type = 'CDWHCL';
Are you able to tell how your Table Check Constraint is represented (with respect to the 'br') in your Database and Designer?
Regards,
Tony

Ideas on how to export from Oracle (11g) for import into a MSSQL server?

Hi folks,
Crazy question yes...but I have some devs that are needing to export some schemas/tables from Oracle to somehow import into MSSQL.
Are there any utilities our there, hopefully from Oracle that would allow for this? So far, all I can think of are to generate the DDL from the Oracle instance, and then do some scripts to output the data in a cvs type format....
Anything better than this?
Any links or suggestions greatly appreciated.
Thank you in advance,
cayenne

cayenne wrote:
EdStevens wrote:
cayenne wrote:
sybrand_b wrote:
Heterogeneous Services using ODBC or OLEDB.
No need for stupid csv files.
Sybrand Bakker
Senior Oracle DBAThank you for the prompt reply. I'm looking into this and am finding some Oracle documentation, but just off the top of my head, this approach seems to assume the 2 database servers can see or talk to each other.
What if the oracle box is on one network...that cannot connect to the network that the MSSQL box is on. For example..running off some test data from one company to another company with networks that do not cross?Assuming some acceptable export/import utility, how do you plan on getting the file from the source to the target if the two servers are on networks that cannot talk to each other?Well, I'm still trying to gather facts, this is just a question coming in from some developers.
From what it may be, is that the can scp from site to site, but no direct connectivity between the database servers on the sites...isolated, maybe a DMZ...I dunno everything yet.Well, you either have connectivity between the two machines or you don't. Binary decision tree.
IF net_connectivity_exists
     you can use oracle sqlnet enabled capabilities
     OR
     you can transfer an export file via ftp, sftp, scp, etc.
ELSE
     you have to transfer an export file via sneakernet
END-IF

How to run import from Unix Oracle 8.0.5 database with Windows Oracle client?

How to run import from Unix Oracle 8.0.5 database via network with Windows Oracle client? Is it possible? When I try to do it Oracle client just hangs... If not which ones are compatible Aix or Solaris to Unix or it must be a Unix client to connect to Unix Oracle database. Thank's for any help.

Hi,
In our project we are using this type. Since in this project server in UNIX(DEC) and client is running on windows. We have created a listener which always listen requests from client and doing the according to the request.
Step1. create a request table, where u are inserting ur request.
step2. create PRO*C proram, which is listening request on the table. If there is any request, call the imp (executable). This is happening on server side.
Benifit. U can make a request from cleint and ftp the file client sit.
Are u interested in more details and code, pls send a mail to me
---- Boby Jose Thekkanath
[email protected]
Dharma Computers(p) Ltd. Bangalore.
null

Datapump Network Import from Oracle 10.2.0 to 10.1.0

Similar Messages

Maybe you are looking for