10 Defining Replication Schemes

Considering failover and recovery scenarios

As you plan a replication scheme, consider every failover and recovery scenario. For example, subscriber failures generally have no impact on the applications connected to the master databases and can be recovered from without disrupting user service. If a failure occurs on a master database, you should have a means to redirect the application load to a subscriber and continue service with no or minimal interruption. This process is typically handled by a cluster manager or custom software designed to detect failures, redirect users or applications from the failed database to one of its subscribers, and manage recovery of the failed database. See Chapter 12, "Managing Database Failover and Recovery".

When planning failover strategies, consider which subscribers will take on the role of the master and for which users or applications. Also consider recovery factors. For example, a failed master must be able to recover its database from its most up-to-date subscriber, and any subscriber must be able to recover from its master. A bidirectional scheme that replicates the entire database can take advantage of automatic restoration of a failed master. See "Automatic catch-up of a failed master database".

Consider the failure scenario for the unidirectionally replicated database shown in Figure 10-1. In the case of a master failure, the application cannot access the database until it is recovered from the subscriber. You cannot switch the application connection or user load to the subscriber unless you use an ALTER REPLICATION statement to redefine the subscriber database as the master. See "Replacing a master database".

Figure 10-1 Recovering a master in a unidirectional scheme

Description of "Figure 10-1 Recovering a master in a unidirectional scheme"

Figure 10-2 shows a bidirectional distributed workload scheme in which the entire database is replicated. Failover in this type of replication scheme involves shifting the users of the application on the failed database to the application on the surviving database. Upon recovery, the workload can be redistributed to the application on the recovered database.

Figure 10-2 Recovering a master in a distributed workload scheme

Description of "Figure 10-2 Recovering a master in a distributed workload scheme"

Similarly, the users in a split workload scheme must be shifted from the failed database to the surviving database. Because replication in a split workload scheme is not at the database level, you must use an ALTER REPLICATION statement to set a new master database. See "Replacing a master database". Upon recovery, the users can be moved back to the recovered master database.

Propagation replication schemes also require the use of the ALTER REPLICATION statement to set a new master or a new propagator if the master or propagator fails. Higher availability is achieved if two propagators are defined in the replication scheme. See Figure 1-11 for an example of a propagation replication scheme with two propagators.

Making decisions about performance and recovery tradeoffs

When you design a replication scheme, weigh operational efficiencies against the complexities of failover and recovery. Factors that may complicate failover and recovery include the network topology that connects a master with its subscribers and the complexity of the replication scheme. For example, it is easier to recover a master that has been fully replicated to a single subscriber than recover a master that has selected elements replicated to different subscribers.

You can configure replication to work asynchronously (the default), "semi-synchronously" with return receipt service, or fully synchronously with return twosafe service. Selecting a return service provides greater confidence that your data is consistent on the master and subscriber databases. Your decision to use default asynchronous replication or to configure return receipt or return twosafe mode depends on the degree of confidence you require and the performance tradeoff you are willing to make in exchange.

Table 10-1 summarizes the performance and recover tradeoffs of asynchronous replication, return receipt service and return twosafe service.

In addition to the performance and recovery tradeoffs between the two return services, you should also consider the following:

• Return receipt can be used in more configurations, whereas return twosafe can only be used in a bidirectional configuration or an active standby pair.

• Return twosafe allows you to specify a "local action" to be taken on the master database in the event of a timeout or other error encountered when replicating a transaction to the subscriber database.

A transaction is classified as return receipt or return twosafe when the application updates a table that is configured for either return receipt or return twosafe. Once a transaction is classified as either return receipt or return twosafe, it remains so, even if the replication scheme is altered before the transaction completes.

For more information about return services, see "Using a return service".

Distributing workloads

Consider configuring the databases to distribute application workloads and make the best use of a limited number of server machines. For example, it may be efficient and economical to configure the databases in a bidirectional distributed workload replication scheme so that each serves as both master and subscriber, rather than as separate master and subscriber databases. However, a distributed workload scheme works best with applications that primarily read from the databases. Implementing a distributed workload scheme for applications that frequently write to the same elements in a database may diminish performance and require that you implement a solution to prevent or manage update conflicts, as described in Chapter 15, "Resolving Replication Conflicts".

Owner of the replication scheme and tables

The owner and name of the replication scheme and the replicated tables must be identical on both the master and subscriber databases. To ensure that there is a common owner across all databases, you can explicitly specify an owner name with your replication scheme name in the CREATE REPLICATION statement.

For example, to assign an owner named repl to the replication scheme named repscheme, the first line of your CREATE REPLICATION statement would look like:

CREATE REPLICATION rep1.repscheme

If you omit the owner from the name of your replication scheme and the replicated tables, the default owner name, as specified by the login name of the requester or the name set by the UID general connection attribute in the DSN, is used in its place. Your replication scheme does not work if owner names are different across its databases.

Database names

These are the roles of the databases in a replication scheme:

• Master: Applications update the master database. The master sends the updates to the propagator or to the subscribers directly.

• Propagator: The propagator database receives updates from the master database and sends them to subscriber databases.

• Subscriber: Subscribers receive updates from the propagator or the master.

Before you define the replication scheme, you need to define the data source names (DSNs) for the databases in the replication scheme. On Linux or UNIX, create an odbc.ini file. On Windows, use the ODBC Administrator to name the databases and set connection attributes. See "Step 1: Create the DSNs for the master and the subscriber" for an example.

Each database "name" specified in a replication scheme must match the prefix of the database file name without the path specified for the DataStore data store attribute in the DSN definition. A replication scheme that uses the names specified in the Data Source Name data store attributes does not work. To avoid confusion, use the same name for both your DataStore and Data Source Name data store attributes in each DSN definition. For example, if the database path is directory/subdirectory/foo.ds0, then foo is the database name that you should use.

Defining data store elements

To replicate the entire contents of the master database (masterds) to the subscriber database (subscriberds), the ELEMENT description (named ds1) might look like the following:

ELEMENT ds1 DATASTORE
  MASTER masterds ON "system1"
  SUBSCRIBER subscriberds ON "system2"

Identify a database host using the host name returned by the hostname operating system command. Host names containing special characters must be surrounded by double quotes (").

You can choose to exclude certain tables and sequences from the data store element by using the EXCLUDE TABLE and EXCLUDE SEQUENCE clauses of the CREATE REPLICATION statement. When you use the EXCLUDE clauses, the entire database is replicated to all subscribers in the element except for the objects that are specified in the EXCLUDE clauses. Use only one EXCLUDE TABLE and one EXCLUDE SEQUENCE clause in an element description. For example, this element description excludes two tables and one sequence:

ELEMENT ds1 DATASTORE
  MASTER masterds ON "system1"
  SUBSCRIBER subscriberds ON "system2"
  EXCLUDE TABLE tab1, tab2
  EXCLUDE SEQUENCE seq1

You can choose to include only certain tables and sequences in the database by using the INCLUDE TABLE and INCLUDE SEQUENCE clauses of the CREATE REPLICATION statement. When you use the INCLUDE clauses, only the objects that are specified in the INCLUDE clauses are replicated to each subscriber in the element. Use only one INCLUDE TABLE and one INCLUDE SEQUENCE clause in an element description. For example, this element description includes one table and two sequences:

ELEMENT ds1 DATASTORE
  MASTER masterds ON "system1"
  SUBSCRIBER subscriberds ON "system2"
  INCLUDE TABLE tab3
  INCLUDE SEQUENCE seq2, seq3

Defining table elements

To replicate the tab1 and tab2 tables from a master database (named masterds and located on a host named system1) to a subscriber database (named subscriberds on a host named system2), the ELEMENT descriptions (named a and b) might look like the following:

ELEMENT a TABLE tab1
  MASTER masterds ON "system1"
 SUBSCRIBER subscriberds ON "system2"
ELEMENT b TABLE tab2
  MASTER masterds ON "system1"
  SUBSCRIBER subscriberds ON "system2"

For requirements for tables in replication schemes, see "Table requirements and restrictions for replication schemes".

Replicating tables with foreign key relationships

You may choose to replicate all or a subset of tables that have foreign key relationships with one another. However, if the foreign key relationships have been configured with ON DELETE CASCADE, then you must configure replication to replicate all of the tables, either by configuring the replication scheme with a DATASTORE element that does not exclude any of the tables, or by configuring the scheme with a TABLE element for every table that is involved in the relationship.

It is not possible to add a table with a foreign key relationship configured with ON DELETE CASCADE to a pre-existing replication scheme using ALTER REPLICATION. Instead, you must drop the replication scheme, create the new table with the foreign key relationship, and then create a new replication scheme replicating all of the related tables.

Replicating materialized views

A materialized view is a summary of data selected from one or more TimesTen tables, called detail tables. Although you cannot replicate materialized views directly, you can replicate their underlying detail tables in the same manner as you would replicate regular TimesTen tables.

The detail tables on the master and subscriber databases can be referenced by materialized views. However, TimesTen replication verifies only that the replicated detail tables have the same structure on both the master and subscriber. It does not enforce that the materialized views are the same on each database.

If you replicate an entire database containing a materialized or nonmaterialized view as a DATASTORE element, only the detail tables associated with the view are replicated. The view itself is not replicated. A matching view can be defined on the subscriber database, but is not required. If detail tables are replicated, TimesTen automatically updates the corresponding view.

Materialized views defined on replicated tables may result in replication failures or inconsistencies if the materialized view is specified so that overflow or underflow conditions occur when the materialized view is updated.

Replicating sequences

You can use replication to ensure that the current value of a sequence on a subscriber database is always in advance of the current value on the master database, thereby preventing conflicts if the sequence is later used to make updates directly on the subscriber database. For example, you may have an application that uses a sequence to determine primary key values in a replicated table, and a bidirectional configuration that includes a database that must assume the master role when the master database fails. By replicating the sequence, you can guarantee that the same sequence value is not used twice, regardless of which database you update directly.

Sequence replication works by transmitting a new current value from the master database to the subscriber every 20 references to the sequence's NEXTVAL, starting with the first reference. For example, consider a sequence my.seq with a MINVALUE of 1 and an INCREMENT of 2. The very first time that you use my.seq.NEXTVAL in a transaction, the current value of the sequence on the master database is changed to three, and a new current value of 41 is replicated to the subscriber. The next 19 references to my.seq.NEXTVAL on the master database result in no new current value being replicated, since the current value of 41 on the subscriber database is still ahead of the current value on the master. Only on the twenty-first reference to my.seq.NEXTVAL is a new current value, 61, transmitted to the subscriber database, as the subscriber's previous current value of 41 would now be behind the value of 43 on the master.

Sequence replication has these limitations:

• Sequences with the CYCLE attribute cannot be replicated.

• The definition of the replicated sequence on each peer database must be identical.

• No conflict checking is performed on sequences. If you make updates to sequences in both databases in a bidirectional replication configuration without using the RETURN TWOSAFE service, it is possible for both sequences to return the identical NEXTVAL.

If you need to use sequences in a bidirectional replication scheme where updates may occur on either peer, you may instead use a nonreplicated sequence with different MINVALUE and MAXVALUE attributes on each database. For example, you may create sequence my.seq on database DS1 with a MINVALUE of 1 and a MAXVALUE of 100, and the same sequence on DS2 with a MINVALUE of 101 and a MAXVALUE of 200. Then, if you configure DS1 and DS2 with a bidirectional replication scheme, you may make updates to either database using the sequence my.seq with the guarantee that the sequence values never conflict. Be aware that if you are planning on using ttRepAdmin -duplicate to recover from a failure in this configuration, you must drop and then re-create the sequence with a new MINVALUE and MAXVALUE after you have performed the duplicate.

Replicated sequences are intended to be used with replicated tables. Therefore, sequence updates are only replicated when they are followed by or used in updates to replicated tables. Operations on sequences such as SELECT my.seq.NEXTVAL FROM sys.dual, while incrementing the sequence value, are not replicated until they are followed by updates to tables that are replicated. A side effect of this behavior is that these sequence updates are not purged from the log until followed by updates to tables that are replicated. This causes ttRepSubscriberWait and ttRepAdmin -wait to fail when only these sequence updates are present at the end of the log.

To replicate updates to the current value of the seq sequence from a master database (named masterds and located on a host named system1) to a subscriber database (named subscriberds on a host named system2), the element description (named a) might look like the following:

ELEMENT a SEQUENCE seq
  MASTER masterds ON "system1"
  SUBSCRIBER subscriberds ON "system2"

RETURN RECEIPT

TimesTen provides an optional return receipt service to loosely couple or synchronize your application with the replication mechanism.

Specify the RETURN RECEIPT attribute to enable the return receipt service for the subscribers listed in the SUBSCRIBER clause of an element description. With return receipt enabled, when the application commits a transaction for an element on the master database, the application remains blocked until the subscriber acknowledges receipt of the transaction update. If the master is replicating the element to multiple subscribers, the application remains blocked until all of the subscribers have acknowledged receipt of the transaction update.

For example replication schemes that use return receipt services, see Example 10-24 and Example 10-25.

ELEMENT e TABLE tab
    MASTER masterds ON "system1"
    SUBSCRIBER subscriberds ON "system2"
      RETURN RECEIPT

If any of the subscribers are unable to acknowledge receipt of the transaction within a configurable timeout period, the application receives a tt_ErrRepReturnFailed (8170) warning on its commit request. You can use the ttRepXactStatus procedure to check on the status of a return receipt transaction. See "Checking the status of return service transactions" for more information on the return service timeout period.

You can also configure the replication agent to disable the return receipt service after a specific number of timeouts. See "Managing return service timeout errors and replication state changes" for details.

The return receipt service is disabled by default if replication is stopped. See "RETURN SERVICES { ON | OFF } WHEN REPLICATION STOPPED" for details.

RETURN RECEIPT BY REQUEST

RETURN RECEIPT enables notification of receipt for all transactions. You can use RETURN RECEIPT with the BY REQUEST option to enable receipt notification only for specific transactions identified by your application.

If you specify RETURN RECEIPT BY REQUEST for a subscriber, you must use the ttRepSyncSet procedure to enable the return receipt service for a transaction. The call to enable the return receipt service must be part of the transaction (autocommit must be off).

ELEMENT e TABLE tab
    MASTER masterds ON "system1"
    SUBSCRIBER subscriberds ON "system2"
      RETURN RECEIPT BY REQUEST

rc = SQLExecDirect( hstmt, (SQLCHAR *)
    "CALL ttRepSyncSet(0x01, 45, NULL)", SQL_NTS )

If any of the subscribers are unable to acknowledge receipt of the transaction update within a configurable timeout period, the application receives a tt_ErrRepReturnFailed (8170) warning on its commit request. See "Setting the return service timeout period".

You can use ttRepSyncGet to check if a return service is enabled and obtain the timeout value. For example:

Command> CALL ttRepSyncGet();
< 01, 45, 1>
1 row found.

RETURN TWOSAFE BY REQUEST

RETURN TWOSAFE enables notification of commit on the subscriber for all transactions. You can use RETURN TWOSAFE with the BY REQUEST option to enable notification of subscriber commit only for specific transactions identified by the application.

If you specify RETURN TWOSAFE BY REQUEST for a subscriber, you must use the ttRepSyncSet procedure to enable the return twosafe service for a transaction. The call to enable the return twosafe service must be part of the transaction (autocommit must be off).

The ALTER TABLE statement cannot be used to alter a replicated table that is part of a RETURN TWOSAFE BY REQUEST transaction. If DDLCommitBehavior=1, this operation results in error 8051. If DDLCommitBehavior=0, the operation succeeds because a commit is performed before the ALTER TABLE operation, resulting in the ALTER TABLE operation being in a new transaction which is not part of the RETURN TWOSAFE BY REQUEST transaction.

ELEMENT a DATASTORE
    MASTER databaseA ON "system1"
    SUBSCRIBER databaseB ON "system2"
      RETURN TWOSAFE BY REQUEST;

rc = SQLExecDirect( hstmt, (SQLCHAR *)
    "CALL ttRepSyncSet(0x01, 45, 1)", SQL_NTS )

You can use ttRepSyncGet to check if a return service is enabled and obtain the timeout value. For example:

Command> CALL ttRepSyncGet();
< 01, 45, 1>
1 row found.

RETURN TWOSAFE

The return twosafe service ensures that each replicated transaction is committed on the subscriber database before it is committed on the master database. If replication is unable to verify the transaction has been committed on the subscriber, it returns notification of the error. Upon receiving an error, the application can either take a unique action or fall back on preconfigured actions, depending on the type of failure.

The return twosafe service is intended to be used in replication schemes where two databases must stay synchronized. One database has an active role, while the other database has a standby role but must be ready to assume an active role at any moment. Use return twosafe with a bidirectional replication scheme with exactly two databases.

To enable the return twosafe service for the subscriber, specify the RETURN TWOSAFE attribute in the SUBSCRIBER clause in the CREATE REPLICATION or ALTER REPLICATION statement.

ELEMENT a DATASTORE
    MASTER databaseA ON "system1"
    SUBSCRIBER databaseB ON "system2"
      RETURN TWOSAFE

CREATE REPLICATION bidirect
ELEMENT a DATASTORE
    MASTER databaseA ON "system1"
    SUBSCRIBER databaseB ON "system2"
      RETURN TWOSAFE
ELEMENT b DATASTORE
    MASTER databaseB ON "system2"
    SUBSCRIBER databaseA ON "system1"
      RETURN TWOSAFE;

When replication is configured with RETURN TWOSAFE, you must disable autocommit mode

When the application commits a transaction on the master database, the application remains blocked until the subscriber acknowledges it has successfully committed the transaction. Initiating identical updates or deletes on both databases can lead to deadlocks in commits that can be resolved only by stopping the processes.

If the subscriber is unable to acknowledge commit of the transaction update within a configurable timeout period, your application receives a tt_ErrRepReturnFailed (8170) warning on its commit request. See "Setting the return service timeout period".

NO RETURN

Use the NO RETURN attribute to explicitly disable the return receipt or return twosafe service. NO RETURN is the default condition. This attribute is typically set in ALTER REPLICATION statements. See Example 14-13.

Setting the return service timeout period

If your replication scheme is configured with one of the return services described in "Using a return service", a timeout occurs if any of the subscribers are unable to send an acknowledgement back to the master within the time period specified by TIMEOUT.

The default return service timeout period is 10 seconds. You can specify a different return service timeout period by:

• Configuring RETURN WAIT TIME in the CREATE REPLICATION or ALTER REPLICATION statement. A RETURN WAIT TIME of 0 indicates no waiting.

• Calling the ttRepSyncSet procedure with a new returnWait parameter

Once set, the timeout period applies to all subsequent return service transactions until you either reset the timeout period or terminate the application session. The timeout setting applies to all return services for all subscribers.

A return service may time out because of a replication failure or because replication is so far behind that the return service transaction times out before it is replicated. However, unless there is a simultaneous replication failure, failure to obtain a return service confirmation from the subscriber does not mean the transaction has not been or will not be replicated.

You can set other STORE attributes to establish policies that automatically disable return service blocking in the event of excessive timeouts and re-enable return service blocking when conditions improve. See "Managing return service timeout errors and replication state changes".

CREATE REPLICATION bidirect
ELEMENT a DATASTORE
    MASTER databaseA ON "system1"
    SUBSCRIBER databaseB ON "system2"
      RETURN TWOSAFE
ELEMENT b DATASTORE
    MASTER databaseB ON "system2"
    SUBSCRIBER databaseA ON "system1"
      RETURN TWOSAFE
STORE databaseA RETURN WAIT TIME 30
STORE databaseB RETURN WAIT TIME 30;

rc = SQLExecDirect( hstmt, (SQLCHAR *)
    "CALL ttRepSyncSet(NULL, 45, NULL)", SQL_NTS )

Managing return service timeout errors and replication state changes

The replication state can be reset to Stop by a user or by the master replication agent in the event of a subscriber failure. A subscriber may be unable to acknowledge a transaction that makes use of a return service and may time out with respect to the master. If any of the subscribers are unable to acknowledge the transaction update within the timeout period, the application receives an errRepReturnFailed warning on its commit request.

The default return service timeout period is 10 seconds. You can specify a different return service timeout period by:

• Configuring the RETURN WAIT TIME attribute in the STORE clause of the CREATE REPLICATION or ALTER REPLICATION statement

• Calling ttRepSyncSet procedure with a new returnWait parameter

A return service may time out or fail because of a replication failure or because replication is so far behind that the return service transaction times out before it is replicated. However, unless there is a simultaneous replication failure, failure to obtain a return service confirmation from the subscriber does not necessarily mean the transaction has not been or will not be replicated.

This section describes how to detect and respond to timeouts on return service transactions. The main topics are:

• When to manually disable return service blocking

• Establishing return service failure/recovery policies

CREATE REPLICATION myscheme
ELEMENT e TABLE tab
  MASTER masterds ON "server1"
  SUBSCRIBER subscriber1 ON "server2"
  RETURN RECEIPT
  STORE masterds ON "server1"
    RETURN SERVICES ON WHEN REPLICATION STOPPED;

ttRepAdmin -dsn masterds -receiver -name subscriber1 -state stop

ttRepAdmin -dsn masterds -receiver -name subscriber1 -state start

CREATE REPLICATION myscheme
ELEMENT e TABLE tab
  MASTER masterds ON "server1"
  SUBSCRIBER subscriber1 ON "server2",
             subscriber2 ON "server3"
RETURN RECEIPT
STORE masterds ON "server1"
  DISABLE RETURN SUBSCRIBER 5
  RETURN WAIT TIME 30;

CREATE REPLICATION myscheme
ELEMENT e TABLE tab
  MASTER masterds ON "server1"
  SUBSCRIBER subscriber1 ON "server2",
             subscriber2 ON "server3"
RETURN RECEIPT
STORE masterds ON "server1"
  DISABLE RETURN SUBSCRIBER 5
  RESUME RETURN 8;

CREATE REPLICATION myscheme
ELEMENT e TABLE tab
  MASTER masterds ON "server1"
  SUBSCRIBER subscriber ON "server2",
             subscriber2 ON "server3"
RETURN RECEIPT
STORE masterds ON "server1"
  DISABLE RETURN ALL 5
  DURABLE COMMIT ON
  RESUME RETURN 8;

Compressing replicated traffic

If you are replicating over a low-bandwidth network, or if you are replicating massive amounts of data, you can set the COMPRESS TRAFFIC attribute to reduce the amount of bandwidth required for replication. The COMPRESS TRAFFIC attribute compresses the replicated data from the database specified by the STORE parameter in your CREATE REPLICATION or ALTER REPLICATION statement. TimesTen does not compress traffic from other databases.

Although the compression algorithm is optimized for speed, enabling the COMPRESS TRAFFIC attribute has some impact on replication throughput and latency.

CREATE REPLICATION repscheme
ELEMENT d1 DATASTORE
    MASTER dsn1 ON machine1
    SUBSCRIBER dsn2 ON machine2
ELEMENT d2 DATASTORE
    MASTER dsn2 ON machine2
    SUBSCRIBER dsn1 ON machine1
STORE dsn1 ON machine1 COMPRESS TRAFFIC ON;

CREATE REPLICATION scheme
ELEMENT d1 DATASTORE
    MASTER dsn1 ON machine1
    SUBSCRIBER dsn2 ON machine2
ELEMENT d2 DATASTORE
    MASTER dsn2 ON machine2
    SUBSCRIBER dsn1 ON machine1
STORE dsn1 ON machine1 COMPRESS TRAFFIC ON
STORE dsn2 ON machine2 COMPRESS TRAFFIC ON;

Port assignments

If you do not assign a PORT attribute, the TimesTen daemon dynamically selects the port. When ports are assigned dynamically for the replication agents, then the ports of the TimesTen daemons have to match as well. Setting the PORT attribute for one scheme sets it for all schemes.

You must assign static ports if you want to do online upgrades.

If you use CREATE REPLICATION to establish different schemes on the same database with different PORT attributes, TimesTen ignores the setting from the last CREATE REPLICATION statement. In this case, you must use ALTER REPLICATION to change the PORT setting.

When statically assigning ports, it is important to specify the full host name, DSN and port in the STORE attribute of the CREATE REPLICATION statement.

CREATE REPLICATION repscheme
ELEMENT el1 TABLE tab
    MASTER dsn1 ON machine1
    SUBSCRIBER dsn2 ON machine2
ELEMENT el2 TABLE tab
    MASTER dsn2 ON machine2
    SUBSCRIBER dsn1 ON machine1
STORE dsn1 ON machine1 PORT 16080
STORE dsn2 ON machine2 PORT 16083;

Setting the log failure threshold

You can establish a threshold value that, when exceeded, sets an unavailable subscriber to the Failed state before the available log space is exhausted. Use the FAILTHRESHOLD attribute to set the log failure threshold. See Example 10-24.

The default threshold value is 0, which means "no limit." See "Setting connection attributes for logging" for details about log failure threshold values.

If a master sets a subscriber database to the Failed state, it drops all of the data for the failed subscriber from its log and transmits a message to the failed subscriber database. If the master replication agent can communicate with the subscriber replication agent, then the message is transmitted immediately. Otherwise, the message is transmitted when the connection is reestablished. After receiving the message from the master, if the subscriber is configured for bidirectional replication or to propagate updates to other subscribers, it does not transmit any further updates, because its replication state has been compromised.

Any application that connects to the failed subscriber receives a tt_ErrReplicationInvalid (8025) warning indicating that the database has been marked Failed by a replication peer. Once the subscriber database has been informed of its failed status, its state on the master database is changed from Failed to Stop.

Applications can use the ODBC SQLGetInfo function to check if the database it is connected to has been set to the Failed state, as described in"Subscriber failures".

Replicating tables with different definitions

Use the TABLE DEFINITION CHECKING attribute to enable replication of tables that are not identical. For example, if tables have columns in a different order or have a different number of partitions, you can replicate them using this clause. A table is partitioned if columns have been added after its initial creation. See Example 10-18.

Setting the TABLE DEFINITION CHECKING attribute to RELAXED requires that replicated tables have the same key definition, number of columns and column data types. Table definition checking occurs on the subscriber side. Setting this attribute to RELAXED for both master and subscriber has the same effect as setting it for only the subscriber.

The RELAXED setting usually results in slightly slower performance. The change in performance depends on the workload and the number of columns in the tables.

You can set table definition checking to RELAXED temporarily while consolidating tables with multiple partitions and then reset it to EXACT. There is no performance loss for tables with identical structures.

CREATE TABLE t1 (a INT PRIMARY KEY, b INT, c INT);

CREATE TABLE t1 (c INT, a INT PRIMARY KEY, b INT);

CREATE REPLICATION rep1
       ELEMENT e1 TABLE t1
       MASTER dsn1
       SUBSCRIBER dsn2
       STORE dsn2 TABLE DEFINITION CHECKING relaxed;

Command> INSERT INTO t1 VALUES (4,5,6);
1 row inserted.

Command> SELECT * FROM t1;
< 5, 6, 4 >
1 row found.

CREATE TABLE t3 (a INT PRIMARY KEY, b INT);

CREATE TABLE t3 (a INT PRIMARY KEY);

ALTER TABLE t3 ADD COLUMN b INT;

CREATE REPLICATION reppart
       ELEMENT e2 TABLE t3
       MASTER dsn1
       SUBSCRIBER dsn2
       STORE dsn2 TABLE DEFINITION CHECKING RELAXED;

Command> INSERT INTO t3 VALUES (1,2);
1 row inserted.

Command> SELECT * FROM t3;
< 1, 2 >
1 row found.

Single subscriber schemes

The scheme shown in Example 10-21 is a single master and subscriber unidirectional replication scheme. The two databases are located on separate hosts, system1 and system2. We use the RETURN RECEIPT service to confirm that all transactions committed on the tab table in the master database are received by the subscriber.

CREATE REPLICATION repscheme
ELEMENT e TABLE tab
    MASTER masterds ON "system1"
    SUBSCRIBER subscriberds ON "system2"
      RETURN RECEIPT;

CREATE REPLICATION repscheme
ELEMENT e DATASTORE
    MASTER masterds ON "server1"
    SUBSCRIBER subscriberds ON "server2";

Multiple subscriber schemes with return services and a log failure threshold

You can create a replication scheme that includes up to 128 subscriber databases. If you are configuring propagator databases, you can configure up to 128 propagators. Each propagator can have up to 128 subscriber databases. See "Propagation scheme" for an example of a replication scheme with propagator databases.

CREATE REPLICATION twosubscribers
ELEMENT e TABLE tab
    MASTER masterds ON "server1"
    SUBSCRIBER subscriber1ds ON "server2",
               subscriber2ds ON "server3";

CREATE REPLICATION twosubscribers
ELEMENT e TABLE rel.tab
  MASTER masterds ON "server1"
  SUBSCRIBER subscriber1ds ON "server2",
             subscriber2ds ON "server3"
  RETURN RECEIPT
STORE masterds FAILTHRESHOLD 10;

CREATE REPLICATION twosubscribers
ELEMENT e TABLE tab
    MASTER masterds ON "server1"
    SUBSCRIBER subscriber1ds ON "server2"
    SUBSCRIBER subscriber2ds ON "server3" RETURN RECEIPT
STORE masterds FAILTHRESHOLD 10;

CREATE REPLICATION twosubscribers
ELEMENT e TABLE tab
    MASTER masterds ON "server1"
    SUBSCRIBER subscriberds1 ON "server2" RETURN RECEIPT BY REQUEST
    SUBSCRIBER subscriber2ds ON "server3" RETURN RECEIPT
STORE masterds FAILTHRESHOLD 10;

Replicating tables to different subscribers

The replication scheme shown in Example 10-27 establishes a master database, named centralds, that replicates four tables. tab1 and tab2 are replicated to the subscriber backup1ds. tab3 and tab4 are replicated to backup2ds. The master database is located on the finance server. Both subscribers are located on the backupsystem server.

CREATE REPLICATION twobackups
ELEMENT a TABLE tab1
  MASTER centralds ON "finance"
  SUBSCRIBER backup1ds ON "backupsystem"
ELEMENT b TABLE tab2
  MASTER centralds ON "finance"
  SUBSCRIBER backup1ds ON "backupsystem"
ELEMENT d TABLE tab3
  MASTER centralds ON "finance"
  SUBSCRIBER backup2ds ON "backupsystem"
ELEMENT d TABLE tab4
  MASTER centralds ON "finance"
  SUBSCRIBER backup2ds ON "backupsystem";

Propagation scheme

In Example 10-28, the master database sends updates on a table to a propagator that forwards the changes to two subscribers. The master database is centralds on the finance host. The propagator database is propds on the nethandler host. The subscribers are backup1ds on backupsystem1 and backup2ds on backupsystem2.

The replication scheme has two elements. For element a, the changes to the tab table on centralds are replicated to the propds propagator database. For element b, the changes to the tab table received by propds are replicated to the two subscribers, backup1ds and backup2ds.

CREATE REPLICATION propagator
ELEMENT a TABLE tab
  MASTER centralds ON "finance"
  SUBSCRIBER propds ON "nethandler"
ELEMENT b TABLE tab
  PROPAGATOR propds ON "nethandler"
  SUBSCRIBER backup1ds ON "backupsystem1",
             backup2ds ON "backupsystem2";

Bidirectional split workload schemes

In Example 10-29, there are two databases, westds on the westcoast host and eastds on the eastcoast host. Customers are represented in two tables: waccounts contains data for customers in the Western region and eaccounts has data for customers from the Eastern region. The westds database updates the waccounts table and replicates it to the eastds database. The eaccounts table is owned by the eastds database and is replicated to the westds database. The RETURN RECEIPT attribute enables the return receipt service to guarantee that transactions on either master table are received by their subscriber.

CREATE REPLICATION r1
ELEMENT elem_waccounts TABLE waccounts
  MASTER westds ON "westcoast"
  SUBSCRIBER eastds ON "eastcoast" RETURN RECEIPT
ELEMENT elem_eaccounts TABLE eaccounts
  MASTER eastds ON "eastcoast"
  SUBSCRIBER westds ON "westcoast" RETURN RECEIPT;

Bidirectional distributed workload scheme

Example 10-30 shows a bidirectional general workload replication scheme in which the accounts table can be updated on either the eastds or westds database. Each database is both a master and a subscriber for the accounts table.

CREATE REPLICATION r1
ELEMENT elem_accounts_1 TABLE accounts
  MASTER westds ON "westcoast"
  SUBSCRIBER eastds ON "eastcoast"
ELEMENT elem_accounts_2 TABLE accounts
  MASTER eastds ON "eastcoast"
  SUBSCRIBER westds ON "westcoast";

When elements are replicated in this manner, the applications should write to each database in a coordinated manner to avoid simultaneous updates on the same data. To manage update conflicts, include a timestamp column of type BINARY(8) in the replicated table and enable timestamp comparison by including the CHECK CONFLICTS clause in the CREATE REPLICATION statement. See Chapter 15, "Resolving Replication Conflicts" for a complete discussion on how to manage update conflicts.

Example 10-31 shows that the tstamp timestamp column is included in the accounts table. The CREATE REPLICATION statement has been modified to include the CHECK CONFLICTS clause.

CREATE TABLE accounts (custname VARCHAR2(30) NOT NULL,
                       address VARCHAR2(80),
                       curbalance DEC(15,2),
                       tstamp BINARY(8),
                       PRIMARY KEY (custname));

CREATE REPLICATION r1
ELEMENT elem_accounts_1 TABLE accounts
  CHECK CONFLICTS BY ROW TIMESTAMP
    COLUMN tstamp
    UPDATE BY SYSTEM
    ON EXCEPTION ROLLBACK WORK
  MASTER westds ON "westcoast"
  SUBSCRIBER eastds ON "eastcoast"
ELEMENT elem_accounts_2 TABLE accounts
  CHECK CONFLICTS BY ROW TIMESTAMP
    COLUMN tstamp
    UPDATE BY SYSTEM
    ON EXCEPTION ROLLBACK WORK
  MASTER eastds ON "eastcoast"
  SUBSCRIBER westds ON "westcoast";