Set configuration parameters for replication
You must set certain configuration parameters before you start Enterprise Replication. You can set other configuration parameters to customize the behavior of Enterprise Replication.
Parameters to set before you start replication
Set the following configuration parameters in the onconfig file on each database server that you want to include in the replication domain before you start replication:
- DBSERVERNAME specifies the name of the database server. If you use both the DBSERVERNAME and DBSERVERALIASES configuration parameters, set the DBSERVERNAME configuration parameter to the TCP connection and not to a shared-memory connection.
- CDR_DBSPACE specifies the dbspace for the syscdr database.
If not set, the root dbspace is used.
If not set and the database server has a storage pool with sufficient space, the dbspace is configured from the storage pool and the CDR_DBSPACE configuration parameter is set when you define a replication server.
- CDR_QUEUEMEM specifies the maximum amount of memory to be used for the send and receive queues.
- CDR_QDATA_SBSPACE specifies the location of the
row data sbspace.
If not set, you cannot define replication servers.
If not set and the database server has a storage pool with sufficient space, the sbspace is configured from the storage pool and the CDR_QDATA_SBPACE configuration parameter is set when you define a replication server. Otherwise, you cannot define replication servers.
- CDR_SERIAL specifies how to generate non-overlapping (unique) values for serial columns across all database servers in the replication domain.
- CDR_TSINSTANCEID specifies how to generate unique identifiers for time series instances across all database servers in the replication domain.
Logging parameters
By default, if Enterprise Replication detects the potential for a log wrap situation when replication log processing lags behind the current log position, user transactions are blocked. You can configure Enterprise Replication to prevent the blocking of user transactions. Depending on the solutions you need, you might set the following configuration parameters in the onconfig file for each database server:
- CDR_LOG_LAG_ACTION specifies the actions that Enterprise Replication during a potential log wrap situation.
- LOG_STAGING_DIR specifies a directory in which compressed log files are staged.
- CDR_LOG_STAGING_MAXSIZE specifies the maximum size that Enterprise Replication can use to stage log files.
- CDR_MAX_DYNAMIC_LOGS specifies the number of dynamic log file requests that Enterprise Replication can make in one server session.
- DYNAMIC_LOGS specifies that logical logs can be added dynamically.
Encryption parameters
- ENCRYPT_CDR specifies whether to enable encryption. The default
value is
0
, which prevents encryption. - ENCRYPT_CIPHERS specifies which ciphers and cipher modes are used for encryption.
- ENCRYPT_MAC controls the level of Message Authentication Code (MAC) used to ensure message integrity.
- ENCRYPT_MACFILE specifies the full path and file names of the MAC files.
- ENCRYPT_SWITCH specifies the number of minutes between automatic renegotiations of ciphers and keys. (The cipher is the encryption methodology. The secret key is the key that is used to build the encrypted data using the cipher.)
Other parameters
Set the following optional configuration parameters to customize your replication environment:
- CDR_DSLOCKWAIT specifies the number of seconds the data sync component waits for the database locks to be released. When replication is active on an instance, you can increase the amount of time to wait for lock resources to accommodate transactions on replicated tables.
- CDR_SUPPRESS_ATSRISWARN suppresses certain data sync error and warning codes from appearing in ATS and RIS files.
- CDR_DELAY_PURGE_DTC specifies how long to retain rows in delete tables to support the delete wins conflict resolution rule.
- GRIDCOPY_DIR specifies the default directory that is used by the ifx_grid_copy procedure.
- CDR_MAX_FLUSH_SIZE specifies the number of replicated transactions that are applied before the logs are flushed to disk.