Missing Objects Report

The missing objects report provides a general overview of the missing objects referenced in the main objects of your source code.

This is a deprecated version of the SnowConvert documentation, please visit the official site HERE.

What is a "Missing Object"?

Missing object is the term used to refer to missing DDL definitions inside the source code that are being referenced by other objects. The table below shows which objects could be missing objects in each supported language.

Object	Teradata	Oracle	SQL Server	Extensibility Languages
Table	✓	✓	✓	✓
View	✓	✓	✓	✓
Procedure	✓	✓	✓	✓
Function	✓	✓	✓	✓
Macro	✓
Package Function		✓
Package Procedure		✓
*Package		✓
Join Index	✓
Index		✓
Synonym		✓
Database Link		✓
Type	✓	✓	✓
Materialized View		✓

If an asterisk ('*') is listed in the section above, it means that the object is used to call properties from itself that are not considered DDL statements such as constants, variables, or cursors.

Where can I find it?

The missing objects report can be found in a folder named "reports", in the output folder of your conversion. The name of the file itself starts with "MissingObjectReferences" so it can easily be located.

The format of the file is .CSV.

What information does it contain?

The missing objects report contains the following information about all the missing objects found while converting:

Column	Description
PartitionKey	The unique identifier of the conversion.
FileName	The name of the file in which the object is located.
Calling_Object_Type	The type of the object referencing a missing object.
Calling_Object_Database	The database where the object referencing the missing object is deployed. For now, only SQL Server objects can have a database.
Calling_Object_Schema	The schema where the object referencing the missing object is deployed.
Calling_Object_Name	The name of the object referencing the missing object.
Calling_Object_FullName	The full qualified name of the object referencing the missing object.
Referenced_Object_Database	The database where the missing object is deployed. For now, only SQL Server objects can have a database.
Referenced_Object_Schema	The schema where the missing object is deployed.
Referenced_Object_Name	The name of the missing object.
Referenced_Object_FullName	The full qualified name of the missing object.
Line	The line number inside the file where the reference occurs.

Known Issues

Teradata

Variables defined in shell files used in script files like .bteq are considered missing objects because their definition is not part of the input files that SnowConvert processes. For example, the myDB variable is defined in the shell file but this is a file that is not part of the input for SnowConvert. Only the .bteq file will be processed and therefore, line 5 will be marked as a missing reference.

myShellFile.sh

export myDB=exampleDatabase
bteq < example.bteq

example.bteq

```
.LABEL EX_SQE

create multiset volatile table DR as
   select * from ${myDB}.myTable;
```

Oracle

Preprocessing an Oracle workload by splitting packages can result in extra missing references if the package's schema is not specified in the extracted objects.

Original Code

Original Package.sql

CREATE package Schema1.Package1
IS
  CREATE TABLE Table1 (
    col1 INTEGER
  );
  
  CREATE PROCEDURE Proc1
    BEGIN
      SELECT * FROM Schema1.Table1;
    END

END   

Notice that in this case, Table1 is automatically created within the schema Schema1, so the reference in line 9 resolves correctly. However, if a package split process is executed prior to the migration and the resulting files are like these ones:

Modified Code after a package split process

Schema1_Table1.sql

  CREATE TABLE Table1 (
    col1 INTEGER1
  ); 

Schema1_Proc1.sql

CREATE PROCEDURE Proc1
    BEGIN
        SELECT * FROM Schema1.Table1;
    END

The reference on line 3 of the file Schema1_Proc1.sql will be marked as a missing reference, because Table1 was not explicitly created within the schema Schema1.