Example: Parameters for Connecting to the Database in Different Scenarios

NOTE: In the following example, host:port represents a node, where host indicates the name or IP address of the server where the database resides, and port indicates the port number of the server where the database resides.

DR

A customer has two database instances. Database instance A is the production database instance, and database instance B is the DR database instance. When the customer performs a DR switchover, database instance A is demoted to the DR database instance, and database instance B is promoted the production database instance. In this case, to avoid application restart or re-release caused by modifications on the configuration file, the customer can write database instances A and B to the connection string when initializing the configuration file. If the primary database instance cannot be connected, the driver attempts to connect to the DR database instance. For example, database instance A consists of node1, node2, and node3, and database instance B consists of node4, node5, and node6.

The URL can be configured as follows:

jdbc:postgresql://node1,node2,node3,node4,node5,node6/database?priorityServers=3

Load Balancing

A customer has a centralized database instance that consists of one primary node and two standby nodes, that is, node1, node2, and node3. node1 is the primary node, and node2 and node3 are the standby nodes.

If the customer wants to evenly distribute the connections established on the same application to three nodes, the URL can be configured as follows:

jdbc:postgresql://node1,node2,node3/database?loadBalanceHosts=true

CAUTION: When loadBalanceHosts is used, if the connection is established on the standby DN, write operations cannot be performed. If read and write operations are required, do not set this parameter.

Log Diagnosis

If a customer encounters slow data import or some errors that are difficult to analyze, the trace log function can be enabled for diagnosis. The URL can be configured as follows:

jdbc:postgresql://node1/database?loggerLevel=trace&loggerFile=jdbc.log

High Performance

A customer may execute the same SQL statement for multiple times with different input parameters. To improve the execution efficiency, the prepareThreshold parameter can be enabled to avoid repeatedly generating execution plans. The URL can be configured as follows:

jdbc:postgresql://node1/database?prepareThreshold=5

A customer queries 10 million data records at a time. To prevent memory overflow caused by simultaneous return of the data records, the defaultRowFetchSize parameter can be used. The URL can be configured as follows:

jdbc:postgresql://node1/database?defaultRowFetchSize=50000

A customer needs to insert 10 million data records in batches. To improve efficiency, the batchMode parameter can be used. The URL can be configured as follows:

jdbc:postgresql://node1/database?batchMode=true