My Oracle Support Banner

Known Issues With The Database Migration Assistant for Unicode (DMU) Tool version 1.2 (Doc ID 1522722.1)

Last updated on AUGUST 23, 2021

Applies to:

Database Migration Assistant for Unicode - Version 1.2 and later
Information in this document applies to any platform.

Purpose

This is archived content about a previous release of the DMU tool, please see <note 1272374.1> The Database Migration Assistant for Unicode (DMU) Tool for documents about the current version.

To document know issues with the Database Migration Assistant for Unicode (DMU) 1.2 Tool

For general information about the DMU tool please see <note 1272374.1> The Database Migration Assistant for Unicode (DMU) Tool

This note is not intended as a replacement of the DMU documentation or the DMU FAQ but is provided in addition.

* If there are questions about the usage of the DMU tool itself who are not answered in the documentation then please use the DMU OTN discussion forum or log a SR using the product "Database Migration Assistant for Unicode" .

* If there are questions about what action to take on Oracle Applications data or Oracle Applications objects then please see <Note 124721.1> and log an Oracle Applications SR and provide an html export of the scan result using the "with some issues" filter (See point H in this note).

* If there are questions about what action to take on the User data or objects for a non-Oracle Application then please contact the vendor.

Questions and Answers

To view full details, sign in with your My Oracle Support account.

Don't have a My Oracle Support account? Click to get started!


In this Document
Purpose
Questions and Answers
 A) ***Tips to avoid problems when using DMU:***
 A1) Purge the RECYCLEBIN before using / scanning with DMU
 A2) Make sure your database has good statistics
 A3) Check if the compatible parameter is set to your base version (recommendation - not mandatory)
 A4) Check optimizer_features_enable setting.
 A5) If you have materialized views , domain indexes and/or Oracle text indexes (= EBS database)
 A6) Check if there is enough space.
 A7) Make sure that during the conversion phase no application or other client connect to the database.
 A8) Check the source database for unneeded or problematic objects:
 A8.a) Invalid objects.
 A8.b) Orphaned Datapump primary tables
 A8.c) Unneeded sample schema/users.
 A8.d) Leftover Temporary tables using CHAR semantics.
 A8.e) Make sure your database is in good health.
 A9) What about Physical / Logical Standby databases?
 A10) Note the current value of job_queue_processes and aq_tm_processes when using spfile :
 A11) Check the DMU docset and the DMU FAQ
 B) Problems starting DMU or during scanning and actions for datadictionary objects:
 B.1) Starting DMU on windows 64bit gives " unable to create an instance of the Java virtual machine located at path <path>\jvm.dll"
 B.2) Scan result is ok (all is green) but DMU gives DMU-00113: In this release the Database Migration Assistant for Unicode does not support converting databases if they have data needing conversion in data dictionary tables.
 B.3) DMU say's "there are partition keys in the database that contain columns defined with character length semantics" .
 B.4) For some CHAR, VARCHAR2 , LONG or CLOB columns who store binary data (like PDF , doc, docx, jpeg, png , etc files) or encrypted data like hashed/encrypted passwords there are rows who have "Invalid Binary Representation".
 B.5) DMU Failed to install the repository.
 B.6) How to remove "Invalid Binary Representation" or "Need conversion" data in the SESSION_KEY column of SYS.REG$ :
 B.7) How to remove "Invalid Binary Representation" or "Need conversion" in ERR_TEXT of CTXSYS.DR$INDEX_ERROR :
 B.8) How to remove "Invalid Binary Representation" or "Need conversion"  in TEXT column of SYS.ERROR$ :
 B.9) How to remove "Invalid Binary Representation" or "Need conversion" in SYS.TRIGGER$ :
 B.10) How to remove "Invalid Binary Representation" or "Need conversion" in QUERY_TXT Column of SYS.REG_SNAP$ (and SYS.SNAP_REFOP$(SQL_TXT), SYS.SNAP$(QUERY_TXT),  SYS.SUM$(SUMTEXT)) :
 B.11) What is the meaning of the "Flashback data archives exist in database" warning ( with orange bullet) in the migration status panel?
 B.12) starting DMU gives "Bad version number in .class file" error.
  B.13) "calculating the table split threshold " in dmu is very slow (popup in the scan wizard)
 B.14) Ora-01031: Insufficient Privileges when connecting with DMU
 B.15) ORA-29913: error in executing ODCIEXTTABLEOPEN callout (or other errors) on SYS.SYS_TZUV2_AFFECTED_REGIONS during scan
 B.16) ORA-00942: table or view does not exist during scan of the database.
 B.17) DMU marks a yellow triangle besides SYS.SOURCE$ , lists no "Invalid Binary Representation" for SYS.SOURCE$ but a few "Need conversion" and will not convert.
 B.18) How to remove "Invalid Binary Representation" or "Need conversion" in SYS.WRI$_ADV_SQLT_PLANS or SYS.WRI$_ADV_MESSAGE_GROUPS :
 B.19) How to remove "Invalid Binary Representation" or "Need conversion" in DEFAULT$ column of SYS.COL$ :
 B.20) How to remove "Invalid Binary Representation" or "Need conversion" in  SYS.SCHEDULER$_JOB_RUN_DETAILS :
 B.21) why is for some tables the "copy data as create table as select" conversion method not available ?
 B.22) Scan fails with ORA-01426 on some tables who have no rows (size is 0) and are created with deffered segment creation.
 B.23) starting DMU on Microsoft Windows fails with error about a missing MSVCR100.dll
 B.24) starting ./dmu.sh fails with interpreter "/bin/bash" not found file link resolves to "/usr/bin/bash" ksh: ./dmu.sh:  not found ( or similar error)
 C) During conversion phase:
 C.1) When DMU is connected to a RAC db that is NOT started as single instance DMU gives the warning that "the database is not mounted in exclusive mode"
 C.2) ORA-12721 during conversion phase.
 C.3) Various errors and dmu "hanging" during conversion: Check UPFRONT if there is enough enough space in the tablespaces and on OS level.
 D) Errors during the Post Conversion phase:
 D.1) DMU refreshes by default all read only Materialise views (MVIEWS) after conversion.
 D.2) Oracle Text index re-creation fails
 D.3) Oracle EBS ( APPS ) instances : Mview refresh fails with ORA-08002: sequence <name> is not yet defined in this session
 D.4) database alert file shows indexes marked as unusable:
 E) Steps to do after the migration to UTF8 or AL32UTF8 with DMU:
 E1) Purge object statistics kept in Automatic Workload Repository
 E2) Make sure all statistics are updated:
 E3) Use Oracle SQL Developer to verify you data, not toad or sqlplus.
 E4) Check for Invalid objects.
 E5) Set job_queue_processes and aq_tm_processes back to the original value if you use a spfile
 E6) if any of the post conversion actions in Database Properties: Converting tab was set to Generate SQL Script
 F) How to have an idea of what DMU is doing if it "hangs"?
 G) Can I remove/ how to remove the DMU objects after conversion?
 H) How to take an html export of the scan result using the "with some issues" filter:
 I) It's not possible to create the DMU diag package:
References

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.