Enable supplemental logging

Database level supplemental logging is an Oracle requirement that ensures that the Oracle redo log on the source database contains the information required to describe all data changes completely. Enabling supplemental logging is a requirement for all redo log mining replication solutions and not specific to Dbvisit Replicate.

Turning on supplemental logging ensures additional information is inserted into redo stream in order to facilitate replication. The extra level of overhead of adding this information into the redo log is generally less than 1%.

 Supplemental logging is not the default behavior of the Oracle database. It has to be enabled manually after the database is created. It can be enabled at two levels

  1. DATABASE LEVEL
  2. TABLE LEVEL

Supplemental logging is also required in order to ensure the following items are replicated successfully:

  • chained rows
  • clustered tables
  • index organized tables (IOTs)

Supplemental logging can be enabled for

  • For all columns
  • For primary key columns
  • For unique columns
  • For foreign key columns


Dbvisit Replicate automatically turns on supplemental logging and sets the minimal logging level based on the replication requirement. No manual supplemental logging steps are required.

 

Supplemental logging is only required for the source database.


The following SQL is run by Dbvisit Replicate to turn on supplemental logging

SQL> alter database add supplemental log data;


If DDL is enabled, then the following is run

SQL> alter database add supplemental log data (primary key) columns;


For each table to be replicated the following SQL is run

SQL> alter table '||owner||'.'||table_name||' add supplemental log data (primary key) columns;


The supplemental log groups created by Dbvisit Replicate start with a G and then a random hex string: G575B4B0CC2E511E2A20CB8ADB3813

  • For Oracle 10g, 11g and 12c if there is no primary key defined on the table, then Oracle automatically adds additional columns to the supplemental log data to satisfy either the unique key, or if there is no unique key, then it adds all the columns in the table.
  • For Oracle 9i, Oracle does not automatically adds additional columns. The specific supplemental log level has to be defined. Dbvisit Replicate does this automatically.

In Oracle 9i, there is a limit of 33 columns in each supplemental log group. See support.oracle.com note ID 466439.1

So in the case of a table with more than 33 columns you create as many distinct supplemental log groups as required for the same table, e.g., 

ALTER TABLE <table_name> ADD SUPPLEMENTAL LOG GROUP <log_grp1> 
(Column1,....,Column33) ALWAYS; 

ALTER TABLE <table_name> ADD SUPPLEMENTAL LOG GROUP <log_grp2> 
(Column34,....,Column66) ALWAYS; 
 
ALTER TABLE <table_name> ADD SUPPLEMENTAL LOG GROUP <log_grp3> 
(Column67,....,Column99) ALWAYS;

Verify supplemental logging

To check and monitor what supplemental logging has been turned on, the following query can be run

SQL> SELECT supplemental_log_data_min MIN, 
supplemental_log_data_pk PK, 
supplemental_log_data_ui UI, 
supplemental_log_data_fk FK,
supplemental_log_data_all "ALL" 
FROM v$database;

Modifying supplemental log behavior

Please see the variables _ADD_SUPLOG and _MINE_KEY_IMAGE_PREFER_SUPLOG in Internal Variables on the default supplemental log behavior and how to modify this. 

How Dbvisit Replicate works with supplemental logging

Dbvisit Replicate does not use primary key definition directly. It uses all the columns that are supplementally logged by Oracle.
When a table is prepared, Dbvisit Replicate issues: alter table xx.xx add supplemental logging (primary key) columns.
This enables the following Oracle behaviour:
  1. If there is an enabled PK, it will add supplemental logging to only the columns of the PK
  2. If there is no PK, but there is unique key that is guaranteed to be unique (none of the columns are NULL) - use columns of this UK
  3. As last resort, log all columns (except those that can never be logged - LOBs, LONGs, nested tables etc.)
This means that as long as there is a way on how to uniquely identify rows, replication will work as expected. Oracle will supplementally log all the columns needed. 
In certain cases such as filtering, you must ensure that that column is always present in the redo logs.

Here is an example of a conditional user-defined supplemental log group for jobcode and sdate columns where ID is primary key. 

ALTER TABLE AVI.SAMPLE
ADD SUPPLEMENTAL LOG GROUP sample_jobcode (ID,JOBCODE,SDATE) ALWAYS;

The ALWAYS keyword is required to create an unconditional supplemental log group to ensure the column(s) in the filter condition are always logged in the redo log, regardless of whether they appear in the DML statement.