Oracle SQL Developer 3.1 Migration Third Party Databases Issues

Question

Hi,
i had following issues with migrating from db2 v8 to oracle 11.2.
Online:
Due to missing privileges and roles for user db user migrations some steps have failed (CREATE USER -> ORA-01031 ...).
After correcting this like described in "Creating a Database User for the Migration Repository" in sqldev online help this has worked.
The problems are:
a) on the overview page at the end of the migration assistent all steps (CAPTURE, CONVERT, GENERATE, DATAMOVE) are shown as complete, even if nothing has done
b) on page 6/9 into migration assistant all changes for datatype convertion are ignored, for example CHAR to VARCHAR2
c) generated files are not visible, even if you mark refresh on file view
d) after restarting sqldev, generarted files are visible in file view, but when you add generated files to svn error message "svn: File: xxx has inconsitent newlines" is shown
e) after sucessful migration on opened migration project pane "data quality" sourcenumrows are NULL, even if they always NOT NULL and count(*) on any table on both sites are equal
Offline:
Generated skripts contains errors:
./startDump.sh: line 157: syntax error near unexpected token `done'
'/startDump.sh: line 157: `done < "schemas.dat"
Can anybody help?
Thanks in advance
André

Answer 1

Hi kgronau,
thanks for your fast answer.
Today i have found 2 new issue.
When you have opened a migration project from repository, on pane "data quality" sourcenumrows are always null
and
sourcename and targetname shows always databse object names from the database on the first migration project in repository independently of extra section in drop down box model and source.
kgronau wrote:
André,
I used SQL Dev 3.1 and I captured a DB2 database. Then I've changed the rule to map char to varchar2 and started the migration.
When I now check out my custom tables all all of them that had in the source model a char column are now using varchar2.I have tried to changed dataype for target database in place over the drop down box, not the edit rule button. It's a little bit confusing to have this option, when it doesn't works.
After using the edit rule button, all works fine. Just the summary page 9/9 doesn't report changed datatype assignment.
>
Could you please explain what you mean with your option c and d?c)
Yes i'm meaning View -> Files. Sorry but on german windows i have just german menu items. That is sometimes tricky to retranslate for support questions and also not helpful when using the online help where all menu items referenced in english :-(
(Do you have an ideo how can i configure sqldev with english menu aon german windows?)
I think, this problemn is special for output folders under subversion control.
d)
Generated Files after the end of the migration are just visible in output folders under subversion control after restarting sqldev
>
Edited by: kgronau on Mar 7, 2012 12:30 PM
Are you talking about Opening a File viewer window (View -> Files)? In my case I have chosen d:\temp\DB2 as output and monitored it during the migration. It isn't refreshed until I manually click on the refresh button - but once the migration has finished and written the output and when I then click on the refresh button I'll see all the directories and the files included.
Edited by: kgronau on Mar 7, 2012 12:39 PM
When a migration has finished then SQL Developer 3.1 now creates in the top directory an unload_script.sh file which calls the other unload scripts.That's right, all scripts are generated.
Also the data unload scripts were created - I need to find a DB2 on Unix to check the script - a quick check of the windows scripts worked correctly.
Edited by: kgronau on Mar 7, 2012 1:22 PM
These unload shell scripts to unload the data out of a DB2 database are also working.
Unfortunately I'm not able to test the shell script used for a source model unload as my UDB is running on Windows.
Didn't the online source model collection work? For me it looks like it did as you mentioned you changed the char data types to varchar2 and this requires already a connection to the source database - except you used the scripts that were generated using the startDump.sh which has failed. Yes, online source model collection did work. Just the unix shell script produces an error on the source unix system with db2. Please see below th generated script.
So please provide here some more details../startDump.sh was startet for testing purposes without any arguments
./startDump.sh: line 157: syntax error near unexpected token `done'
'/startDump.sh: line 157: `done < "
if [[ $# != 3 ]]; then
echo "Usage: startDump <database> <user> <password>";
exit 1;
fi
ROWTAG="'<row>'";
ENDROWTAG="'</row>'";
COLTAG="'<col><![CDATA['";
ENDCOLTAG="']]></col>'";
# Clear any other dat files
echo "Clearing older data files"
rm -f *.dat
echo "Connnecting to $1 as $2";
db2 -r connect.dat "connect to $1 user $2 using $3";
if [[ $? != 0 ]]; then
echo "Connection failed.";
exit 20;
fi
# GET SCHEMA QUERY.
echo "Get all schemas";
db2 +o -x -r schemas.dat "select SCHEMANAME SCHEMA_NAME from SYSCAT.SCHEMATA WHERE DEFINER <> 'SYSIBM' AND
SCHEMANAME <> 'NULLID' AND SCHEMANAME <> 'SQLJ'
AND SCHEMANAME <> 'SYSTOOLS'";
if [[ $? != 0 ]]; then
echo "Get schemas failed.";
exit 30;
fi
# Loop through file containing schema names and extract db objects for each of them
while read SCHEMA_NAME
do
# Create schema directory
rm -rf "${SCHEMA_NAME}";
mkdir "${SCHEMA_NAME}";
if [[ $? != 0 ]]; then
echo "Could not create schema directory ${SCHEMA_NAME}.";
exit 40;
fi
echo "Get all tables for schema $SCHEMA_NAME";
tablesFile="${SCHEMA_NAME}/""tables.dat";
# GET TABLES QUERY. */
db2 -x +o -r $tablesFile "select "$ROWTAG", "$COLTAG"||COLUMNS.TABSCHEMA||"$ENDCOLTAG", "$COLTAG"||COLUMNS.TABNAME||"$ENDCOLTAG",
"$COLTAG"||COLUMNS.COLNAME||"$ENDCOLTAG", "$COLTAG"||(CASE WHEN (COLUMNS.CODEPAGE = 0 and (COLUMNS.TYPENAME = 'VARCHAR' OR COLUMNS.TYPENAME = 'CHAR'
OR COLUMNS.TYPENAME = 'LONG VARCHAR' OR COLUMNS.TYPENAME = 'CHARACTER')) THEN COLUMNS.TYPENAME || ' FOR BIT DATA'
ELSE COLUMNS.TYPENAME END)||"$ENDCOLTAG", "$COLTAG"||CHAR(COLUMNS.LENGTH)||"$ENDCOLTAG",
"$COLTAG"||CHAR(COLUMNS.SCALE)||"$ENDCOLTAG", "$COLTAG"||COLUMNS.NULLS||"$ENDCOLTAG",
"$COLTAG"||COALESCE(COLUMNS.DEFAULT, '')||"$ENDCOLTAG", "$ENDROWTAG" from
SYSCAT.COLUMNS COLUMNS, SYSCAT.TABLES TABLES WHERE
COLUMNS.TABSCHEMA = '${SCHEMA_NAME}' AND
COLUMNS.TABNAME = TABLES.TABNAME AND
COLUMNS.TABSCHEMA = TABLES.TABSCHEMA AND
TABLES.TYPE = 'T'
ORDER BY COLUMNS.TABNAME, COLUMNS.COLNO";
if [[ $? != 0 ]]; then
echo "No tables found.";
fi
# GET SYNONYMS QUERY. */
echo "Get all synonyms for schema $SCHEMA_NAME";
synonymsFile="${SCHEMA_NAME}/""synonyms.dat";
db2 -x +o -r $synonymsFile "select "$ROWTAG", "$COLTAG"||TABNAME||"$ENDCOLTAG", "$COLTAG"||BASE_TABSCHEMA||"$ENDCOLTAG",
"$COLTAG"||BASE_TABNAME||"$ENDCOLTAG", "$ENDROWTAG" from syscat.tables
where tabschema = '${SCHEMA_NAME}' and type = 'A'";
if [[ $? != 0 ]]; then
echo "No synonyms found.";
fi
# GET VIEW QUERY. */
echo "Get all views for schema $SCHEMA_NAME";
viewsFile="${SCHEMA_NAME}/""views.dat";
db2 -x +o -r $viewsFile "select "$ROWTAG", "$COLTAG"||VIEWSCHEMA||"$ENDCOLTAG", "$COLTAG"||VIEWNAME||"$ENDCOLTAG",
"$COLTAG"||COALESCE(TEXT, '')||"$ENDCOLTAG",
"$COLTAG"||DEFINER||"$ENDCOLTAG", "$COLTAG"||READONLY||"$ENDCOLTAG", "$COLTAG"||VALID||"$ENDCOLTAG", "$ENDROWTAG"
from syscat.views
WHERE VIEWSCHEMA = '${SCHEMA_NAME}'
ORDER BY VIEWNAME";
if [[ $? != 0 ]]; then
echo "No views found.";
fi
# GET INDEXES QUERY. */
echo "Get all indexes for schema $SCHEMA_NAME";
indexesFile="${SCHEMA_NAME}/""indexes.dat";
db2 -x +o -r $indexesFile "select "$ROWTAG", "$COLTAG"||INDSCHEMA||"$ENDCOLTAG", "$COLTAG"||INDNAME||"$ENDCOLTAG",
"$COLTAG"||TABSCHEMA||"$ENDCOLTAG", "$COLTAG"||TABNAME||"$ENDCOLTAG", "$COLTAG"||INDEXTYPE||"$ENDCOLTAG",
"$COLTAG"||UNIQUERULE||"$ENDCOLTAG", "$ENDROWTAG" from SYSCAT.INDEXES
WHERE INDSCHEMA = '${SCHEMA_NAME}' AND UNIQUERULE <> 'P'
ORDER BY TABNAME, INDNAME";
if [[ $? != 0 ]]; then
echo "No indexes found.";
fi
# GET INDEX DETAILS QUERY. */
echo "Get all index details for schema $SCHEMA_NAME";
indexeDetailsFile="${SCHEMA_NAME}/""indexDetails.dat";
db2 -x +o -r $indexeDetailsFile "select "$ROWTAG", "$COLTAG"||INDSCHEMA||"$ENDCOLTAG", "$COLTAG"||INDNAME||"$ENDCOLTAG",
"$COLTAG"||COLNAME||"$ENDCOLTAG", "$COLTAG"||CHAR(COLSEQ)||"$ENDCOLTAG", "$ENDROWTAG" from SYSCAT.INDEXCOLUSE
WHERE INDSCHEMA = '${SCHEMA_NAME}'";
if [[ $? != 0 ]]; then
echo "No index details found.";
fi
# GET TRIGGERS QUERY. */
echo "Get all triggers for schema $SCHEMA_NAME";
triggersFile="${SCHEMA_NAME}/""triggers.dat";
db2 -x +o -r $triggersFile "select "$ROWTAG", "$COLTAG"||TRIGSCHEMA||"$ENDCOLTAG",
"$COLTAG"||TRIGNAME||"$ENDCOLTAG", "$COLTAG"||DEFINER||"$ENDCOLTAG", "$COLTAG"||TABSCHEMA||"$ENDCOLTAG",
"$COLTAG"||TABNAME||"$ENDCOLTAG", "$COLTAG"||TRIGEVENT||"$ENDCOLTAG", "$COLTAG"||VALID||"$ENDCOLTAG",
"$COLTAG"||COALESCE(TEXT, '')||"$ENDCOLTAG",
"$COLTAG"||COALESCE(REMARKS, '')||"$ENDCOLTAG", "$ENDROWTAG"
from SYSCAT.TRIGGERS
WHERE TRIGSCHEMA = '${SCHEMA_NAME}'";
if [[ $? != 0 ]]; then
echo "No triggers found.";
fi
# The for GET Promary Key CONSTRAINT QUERY. */
echo "Get all primary keys for schema $SCHEMA_NAME";
primarykeysFile="${SCHEMA_NAME}/""primarykeys.dat";
db2 -x +o -r $primarykeysFile "select "$ROWTAG", "$COLTAG"||X.CONSTNAME||"$ENDCOLTAG", "$COLTAG"||X.TYPE||"$ENDCOLTAG",
"$COLTAG"||X.TABSCHEMA||"$ENDCOLTAG", "$COLTAG"||X.TABNAME||"$ENDCOLTAG", "$COLTAG"||Z.COLNAME||"$ENDCOLTAG",
"$COLTAG"||CHAR(Z.COLSEQ)||"$ENDCOLTAG", "$COLTAG"||COALESCE(X.REMARKS, '')||"$ENDCOLTAG", "$ENDROWTAG" from
(select CONSTNAME, TYPE, TABSCHEMA, TABNAME, REMARKS from SYSCAT.TABCONST where (type = 'P' OR type = 'U')) X
FULL OUTER JOIN
(select COLNAME, COLSEQ, CONSTNAME, TABSCHEMA, TABNAME from SYSCAT.KEYCOLUSE) Z
on
(X.CONSTNAME = Z.CONSTNAME and X.TABSCHEMA = Z.TABSCHEMA and X.TABNAME = Z.TABNAME)
WHERE X.TABSCHEMA='${SCHEMA_NAME}'
ORDER BY X.CONSTNAME";
if [[ $? != 0 ]]; then
echo "No primary keys found.";
fi
# The for GET Check constraints QUERY. */
echo "Get all Check constraints for schema $SCHEMA_NAME";
constraintsFile="${SCHEMA_NAME}/""checkConstraints.dat";
db2 -x +o -r $constraintsFile "SELECT "$ROWTAG", "$COLTAG"||A.CONSTNAME||"$ENDCOLTAG", "$COLTAG"|| COALESCE(TEXT, '') ||"$ENDCOLTAG", "$COLTAG"|| A.TABSCHEMA||"$ENDCOLTAG", "$COLTAG"|| A.TABNAME ||"$ENDCOLTAG", "$COLTAG"|| COLNAME ||"$ENDCOLTAG", "$ENDROWTAG" FROM SYSCAT.CHECKS A , SYSCAT.COLCHECKS B
WHERE A.CONSTNAME = B.CONSTNAME AND A.TABSCHEMA = B.TABSCHEMA AND A.TABNAME=B.TABNAME AND A.TABSCHEMA = '${SCHEMA_NAME}'";
if [[ $? != 0 ]]; then
echo "No check constraints found.";
fi
done < "schemas.dat"
# GET PROCEDURES QUERY. */
. getProcedures.sh schemas.dat
# The for GET Foreign Key CONSTRAINT QUERY. */
. getForeignKeys.sh schemas.dat