Page tree
Skip to end of metadata
Go to start of metadata

This section deals with issues that can occur when you analyze a DB2 for z/OS extraction. Some messages are related to information stored in data files. To understand the problem and to check the content of these files, you can use the SQL query specified in the extractor source code in steps STEP03 to STEP24 (see CAST and zOS for more information).

Messages related to Manifest file

"Current manifest file contains a missing entry: <entry> is required."

The analyzer looks for an entry in the Manifest file and does not find it. As a consequence it cannot process the extraction files correctly. These entries are specified in the "GENERATED FILES" and "EXTRACTED COLUMNS" sections.

Some of them are mandatory:

  • GLOBAL.DATABASES
  • GLOBAL.SCHEMA
  • GLOBAL.DTP
  • TABLES.DEF

Check the Manifest file and add these entries in both sections. You should have a look to the extraction JCL to find the default settings for these entries.

"Incorrect entry format at line (<lineno>): actual columns = <colon> , expected columns =<colon>."

In the "EXTRACTED COLUMNS" section, an entry has been altered and some of expected columns are missing. The analyzer cannot process the associated data file correctly.
You should have a look to the extraction JCL to find the default settings for these entries.

Messages related to file format

"'< file>' DB2 for z/OS extracted file doesn't match expected format."

This error occurs in the z/OS offline analysis mode. It means that the format of the file "<file>" is not correct. This is generally due to a problem of coherence between a "size" column and its associated "data" column.
In this case, you should open this file and check if each "size" column contains the size of its associated "data" column.

If you found a mistake, then you must change the configuration of the extractor for the step that generates the file "<file>". Then, you can either run the extractor again or you can change the value manually for each line of the file.

You should also verify if the utility used to transfer files from host to Windows does not remove the right-most spaces of files. In this case, the utility must be configured to keep file padding.

Messages related to invalid objects

"System datatype '' located in schema '<schname>' and specified at line (<lineno>) is not valid."

You should open the file associated to the entry GLOBAL.DTP to verify the content of the line mentioned in the message.
You can remove these lines from the file before to rerun the analysis. However, it could be interesting to speak about this problem with the DBA.

"Table '' in schema 'A' on database 'A0PDB' specified at line (1) is not valid."

You should open the file associated to the entry TABLES.DEF to verify the content of the line mentioned in the message.

You can remove these lines from the file before to rerun the analysis. However, it could be interesting to speak about this problem with the DBA.

Messages related to column definition

"Column '<colname>' (with datatype name = '') that belongs to table '<tabname>' in schema '<schname>' specified at line (<lineno>) is not valid."

It is possible that the data type name specified in a column definition (see file associated to entry TABLES.COL) is blanked. Normally the extractor replaces it by the COLTYP field. If not, then you should ask the DBA why the TYPNAM field is empty and how to set a value in this field.
You can either modify the column file manually or change the query used in step STEP11 of the JCL and rerun the extraction process.

Messages related to indexes and primary keys

"Index '<idname>' defined on table '<tabname>' (schema= '<schname>') is not applied to any columns: please check columns definition file."

"Wrong entry specified in indexes definition file at line (<lineno>):
Index '<idname>' is defined on table '<tabname>' (schema = '<schname>') that does not exist."

The index references a table that has not been saved in the KB. Look at the related data in files associated to entries GLOBAL.IDX, GLOBAL.IDXCOL, TABLES.DEF and GLOBAL.SCHEMAS.

It is possible that this error is a consequence of the filters used to analyze the DB2 extraction. You should look for log file messages that could mention these objects. Some of these messages are shown below.

"Wrong entry specified in table dependencies file at line (<lineno>):
Table '<tabname>' in schema '<schname>' to which primary key '<pkname>' belongs to does not exist."

The primary key references a table that has not been saved in the KB. Look at the related data in files associated to entries TABLES.DEF, TABLES.DEP and GLOBAL.SCHEMAS.

It is possible that this error is a consequence of the filters used to analyze the DB2 extraction. You should look for log file messages that could mention these objects. Some of these messages are shown below.

"A primary key '<pkname>' (on table '<tabname>' in schema '<schname>') is defined on column '<colname>' that does not exist: please check column definition file."

This message can be the consequence of several situations:

  • No column has been saved for the table
  • A column of a table references another column belonging to another table (via primary key or foreign key) that has not been saved

Look at the related data in files associated to entries TABLES.DEF, TABLES.DEP, TABLES.COL and GLOBAL.SCHEMAS.It is possible that this error is a consequence of the filters used to analyze the DB2 extraction. You should look for log file messages that could mention these objects. Some of these messages are shown below.

Messages related filters

When you use a filter file to analyze a DB2 for z/OS extraction, the analyzer sends messages in the log file to indicate that an object is rejected. These messages can be used to understand other troubles involving missing objects.
These messages look like:
"Database name '<dbname>' specified at line (<lineno>) was rejected by the active filter file."


  • No labels