

Configuration

The pool configuration is represented by ConnectionPoolConfig instance. Instances of this class can be created using its companion object's apply method. Every parameter of this method has some default value so you can use named parameters to set only some configuration options and keep others at defaults.

Options

The list below contains available configuration options.

• name

  Name of the pool — appears in logs.

  Default value: "unnamed".

---

• size

  Size of the pool. The pool will maintain at most size connections at any given time, this includes both idle and in-use connections. If some connection is closed the pool will do its best to replace it as soon as possible so that the pool always contains size connections.

  Default value: 20.

---

• connectionValidateTimeout

  Connections are always validated before being returned to clients. If the connection validity can't be determined in this time, the connection is considered invalid and is closed and eventually removed from the pool.

  Default value: Timeout(5.seconds).

---

• connectionCreateTimeout

  If opening the connection doesn't finish in this time, the attempt is aborted. Note that this property doesn't control the maximum time which clients will wait for connections — it's for pool's internal connection requests.

  Default value: Timeout(5.seconds).

---

• connectionRollbackTimeout

  When clients return connections to the pool, the pool rolls back any transaction that may be in progress so that other clients are unaffected by previous connection usage. This property controls maximum time that the rollback operation can take. If it takes longer the connection is closed and removed from the pool.

  Default value: Timeout(5.seconds).

---

• taskSchedulerFactory

  A function that creates task scheduler for pool's internal use. The scheduler is used to schedule connection request timeouts.

  Default value: A function creating JDK scheduler that uses a single thread and the global execution context for internal asynchronous executions.

  () => new JdkScheduler(Executors.newSingleThreadScheduledExecutor())
                        (ExecutionContext.global)
  

---

• executionContext

  An execution context that will be used by the pool internally to execute callbacks on Futures and any asynchronous operations it needs. The operations executed are guaranteed to be non-blocking.

  Default value: ExecutionContext.global.

Examples

The snippet below creates a pool specifying its name, size and validation timeout.

import scala.concurrent.duration._

import io.rdbc.sapi._
import io.rdbc.pool.sapi._

val cf: ConnectionFactory = ???

val pool = ConnectionPool(cf, ConnectionPoolConfig(
              name = "mypool",
              size = 50,
              connectionValidateTimeout = 5.seconds.timeout)
           )