Skip to content
This repository has been archived by the owner on Jan 17, 2024. It is now read-only.

gamlerhart/adbcj

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Asynchronous Database Connectivity in Java (ADBCJ)

ADBCJ allows you to access a relational database in a asynchronous, non-blocking fashion. The API is inspired by JDBC, but makes all calls asynchronous.

The asynchronous access prevents any blocked threads, which just wait for the result of the database.

It also allows to pipeline operations, which are independent. Depending on the application, this can give a significant performance gain.

ADBCJ is intended as low level foundation. Therefore it's is written in Java, so other languages like Scala, Groovy, Kotlin etc can cosume it to

Getting Started

All the code of this section is in the 'adbcj-demo' directory.

Add the dependencies

First add the Maven dependency and repository. Currently only snapshots are available.

<repository>
    <snapshots>
        <enabled>false</enabled>
    </snapshots>
    <id>central</id>
    <name>bintray</name>
    <url>https://jcenter.bintray.com</url>
</repository>

Add you're dependencies now. For MySQL:

<dependency>
    <groupId>org.adbcj</groupId>
    <artifactId>adbcj-api</artifactId>
    <version>0.9</version>
</dependency>
<dependency>
    <groupId>org.adbcj</groupId>
    <artifactId>mysql-async-driver</artifactId>
    <version>0.9</version>
</dependency>

Or for H2:

<dependency>
    <groupId>org.adbcj</groupId>
    <artifactId>adbcj-api</artifactId>
    <version>0.9</version>
</dependency>
<dependency>
    <groupId>org.adbcj</groupId>
    <artifactId>h2-async-driver</artifactId>
    <version>0.9</version>
</dependency>

Connect to Database

First we create a connection manager, which holds all connections and other resources, like thread pools. Usually you have one connection manager in an application.

The connection URL have a similar format to JDBC connection strings. For Mysql:

adbcj:mysql://localhost/the-database

For H2:

adbcj:h2://localhost/the-database

An H2 example. Let's create the connection manager

final ConnectionManager connectionManager = ConnectionManagerProvider.createConnectionManager(
        "adbcj:h2://localhost/the-datanbase",
        "user",
        "password"
);

Then, connect.

try {
    Connection connection = connectionFuture.get();
    System.out.println("Connected!");
    connection.close().get();
    System.out.println("Closed!");
    connectionManager.close();
} catch (InterruptedException e) {
    Thread.currentThread().interrupt();
} catch (ExecutionException e) {
    e.printStackTrace();
    System.exit(-1);
}

Go Async

ADBCJ is built for asynchronous operations. If you want synchronous operations, use JDBC.

So, when we just do a blocking Future.get(), there is no use in ADBCJ. For direct ADBCJ programming, operations return Java 8 completable future. Therefore, you can take advantage of all the completable future features to compose your async operations

connectionManager.connect().thenCompose(connection -> {
    System.out.println("Connected!");
    return connection.close();
}).thenCompose(closeComplete -> {
    System.out.println("Close complete!");
    return connectionManager.close();
}).whenComplete((complete, error) -> {
    if (error != null) {
        error.printStackTrace();
    }
    System.exit(-1);
});

In case Java 8 CompletableFutures don't fit you're needs, every ADBCJ operation also takes a low level DbCallback. This allows you to plug it into your own futures or async handling style. The DbCallback, on success is called with the value or null. On failure it is called with the exception.

connectionManager.connect((connection, connectionFailure) -> {
    if (connectionFailure == null) {
        System.out.println("Connected!");
        // No failure, continue the operations
        connection.close((connectionClosed, closeFailure) -> {
            if (closeFailure != null) {
                closeFailure.printStackTrace();
            } else {
                System.out.println("Closed!");
            }

            // At the end, close the connection manger
            connectionManager.close((managerClosed, managerCloseFailure) -> {
                if (managerCloseFailure != null) {
                    managerCloseFailure.printStackTrace();
                }
                System.exit(0);
            });
        });
    } else {
        // Otherwise, back out and print the error
        connectionFailure.printStackTrace();
        System.exit(-1);
    }
});

The first SQL statements

ADBCJ has executeUpdate, executeQuery, prepareUpdate, prepareQuery etc similar to JDBC. So you can use these operations to update the database.

ADBCJ tries it's best to send any query immediately to the database when possible. It tries the best to avoid waiting on any round trips. So, if you're queries and statements do not depend on each other, send them right away. Then wait for the result of all sent queries.

connectionManager.connect().thenAccept(connection -> {
    CompletableFuture<Result> create = connection.executeUpdate(
            "CREATE TABLE IF NOT EXISTS posts(\n" +
                    "  id int NOT NULL AUTO_INCREMENT,\n" +
                    "  title varchar(255) NOT NULL,\n" +
                    "  content TEXT NOT NULL,\n" +
                    "  PRIMARY KEY (id)\n" +
                    ");");
    // NOTE: As long operation does not depend on the previous one, you can issue then right away.
    // ADBCJ will immediately send these requests to the database, when possible
    CompletableFuture<Result> firstInsert = connection.executeUpdate("INSERT INTO posts(title,content) VALUES('The Title','TheContent')");
    CompletableFuture<Result> secondInsert = connection.executeUpdate("INSERT INTO posts(title,content) VALUES('Another Title','More Content')");

    // Once you need to wait for previous requests, the Java 8 completable futures will help you
    CompletableFuture<Void> createAndInsertDone = CompletableFuture.allOf(create, firstInsert, secondInsert);

    // And then, a simple query
    createAndInsertDone.thenCompose(res -> {
        return connection.executeQuery("SELECT * FROM posts");
    }).thenAccept(queryResult -> {
        // NOTE: ADBCJ default result sets are regular Java collections.
        // They start at index 0 (ZERO)
        // And interate like regular collections
        for (Row row : queryResult) {
            System.out.println("ID: " + row.get("ID").getLong() + " with title " + row.get("title").getString());
        }
    }).whenComplete((res, failure) -> {
        if (failure != null) {
            failure.printStackTrace();
        }
        connection.close();
        connectionManager.close();
    });
});

Dealing with Failures

Let's create a error in our SQL. Example:

private static void oupsError(ConnectionManager connectionManager) {
    connectionManager.connect().thenAccept(connection -> {
        connection.executeQuery(
                "SELECT * FROM not-existing-table")
                .thenAccept(queryResult -> {
                    // No result, since wrong SQL.
                }).whenComplete((res, failure) -> {
            if (failure != null) {
                failure.printStackTrace();
            }
            connection.close();
            connectionManager.close();
        });
    });
}

This leads to this exception or similar stack trace. Notice something? The stack trace has no where the stack of our application in it. Nothing points us at the origin of the error in the oupsError method. We only have internal stack traces.

This is a problem from the asynchronous nature of the ADBCJ: The handling of the connection is done on a event loop of the driver.

java.util.concurrent.CompletionException: org.adbcj.h2.H2DbException: Syntax error in SQL statement "SELECT * FROM NOT[*]-EXISTING-TABLE "; expected "identifier"
    at java.util.concurrent.CompletableFuture.encodeThrowable(CompletableFuture.java:292)
    at java.util.concurrent.CompletableFuture.completeThrowable(CompletableFuture.java:308)
    at java.util.concurrent.CompletableFuture.uniAccept(CompletableFuture.java:647)
    at java.util.concurrent.CompletableFuture$UniAccept.tryFire(CompletableFuture.java:632)
    at java.util.concurrent.CompletableFuture.postComplete(CompletableFuture.java:474)
    at java.util.concurrent.CompletableFuture.completeExceptionally(CompletableFuture.java:1977)
    at org.adbcj.support.DbCompletableFuture.onComplete(DbCompletableFuture.java:21)
    at org.adbcj.h2.decoding.StatementPrepare.requestFailedContinue(StatementPrepare.java:48)
    at org.adbcj.h2.decoding.StatusReadingDecoder.handleException(StatusReadingDecoder.java:55)
    at org.adbcj.h2.decoding.StatementPrepare.handleException(StatementPrepare.java:39)
    at org.adbcj.h2.RequestCreator$ContinueSql.handleException(RequestCreator.java:238)
    at org.adbcj.h2.decoding.AnswerNextRequest.handleException(AnswerNextRequest.java:28)
    at org.adbcj.h2.decoding.StatusReadingDecoder.decode(StatusReadingDecoder.java:36)
    at org.adbcj.h2.decoding.Decoder.decode(Decoder.java:31)
    at io.netty.handler.codec.ByteToMessageDecoder.callDecode(ByteToMessageDecoder.java:411)
    at io.netty.handler.codec.ByteToMessageDecoder.channelRead(ByteToMessageDecoder.java:248)
    at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:373)
    at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:359)
    at io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:351)
    at io.netty.channel.DefaultChannelPipeline$HeadContext.channelRead(DefaultChannelPipeline.java:1334)
    at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:373)
    at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:359)
    at io.netty.channel.DefaultChannelPipeline.fireChannelRead(DefaultChannelPipeline.java:926)
    at io.netty.channel.nio.AbstractNioByteChannel$NioByteUnsafe.read(AbstractNioByteChannel.java:129)
    at io.netty.channel.nio.NioEventLoop.processSelectedKey(NioEventLoop.java:651)
    at io.netty.channel.nio.NioEventLoop.processSelectedKeysOptimized(NioEventLoop.java:574)
    at io.netty.channel.nio.NioEventLoop.processSelectedKeys(NioEventLoop.java:488)
    at io.netty.channel.nio.NioEventLoop.run(NioEventLoop.java:450)
    at io.netty.util.concurrent.SingleThreadEventExecutor$5.run(SingleThreadEventExecutor.java:873)
    at io.netty.util.concurrent.DefaultThreadFactory$DefaultRunnableDecorator.run(DefaultThreadFactory.java:144)
    at java.lang.Thread.run(Thread.java:745)
Caused by: org.adbcj.h2.H2DbException: Syntax error in SQL statement "SELECT * FROM NOT[*]-EXISTING-TABLE "; expected "identifier"
    at org.adbcj.h2.H2DbException.create(H2DbException.java:38)
    at org.adbcj.h2.decoding.StatusReadingDecoder.decode(StatusReadingDecoder.java:37)
    ... 18 more

Luckily, ADBCJ has a debug option to capture the entry point of every operation. Add -Dorg.adbcj.debug=true as a JVM start up parameter. This trigger ADBCJ to capture all entry points into the driver and add it to errors. This is a development and debug feature. It has a high overhead. As an alternative to the JVM flag, you can it on a per connection basis, by adding "org.adbcj.debug"="true" to the properties map when connecting.

The stack trace now containes the origin of the error. Do you spot 'TutorialDealingWithErrors.oupsError' in the second stack trace? That's the true origin where the error started.

java.util.concurrent.CompletionException: org.adbcj.h2.H2DbException: Syntax error in SQL statement "SELECT * FROM NOT[*]-EXISTING-TABLE "; expected "identifier"
    at java.util.concurrent.CompletableFuture.encodeThrowable(CompletableFuture.java:292)
    at java.util.concurrent.CompletableFuture.completeThrowable(CompletableFuture.java:308)
    at java.util.concurrent.CompletableFuture.uniAccept(CompletableFuture.java:647)
    at java.util.concurrent.CompletableFuture$UniAccept.tryFire(CompletableFuture.java:632)
    // SNIP
Caused by: org.adbcj.h2.H2DbException: Syntax error in SQL statement "SELECT * FROM NOT[*]-EXISTING-TABLE "; expected "identifier"
    at java.lang.Thread.getStackTrace(Thread.java:1556)
    at org.adbcj.support.stacktracing.StackTracingOptions$1.captureStacktraceAtEntryPoint(StackTracingOptions.java:16)
    at org.adbcj.support.AbstractConnectionManager.entryPointStack(AbstractConnectionManager.java:102)
    at org.adbcj.h2.H2ConnectionManager.connect(H2ConnectionManager.java:69)
    at org.adbcj.h2.H2ConnectionManager.connect(H2ConnectionManager.java:60)
    at org.adbcj.ConnectionManager.connect(ConnectionManager.java:48)
    at info.adbcj.demo.TutorialDealingWithErrors.oupsError(TutorialDealingWithErrors.java:33)
    at info.adbcj.demo.TutorialDealingWithErrors.main(TutorialDealingWithErrors.java:19)
Caused by: org.adbcj.DbException: Syntax error in SQL statement "SELECT * FROM NOT[*]-EXISTING-TABLE "; expected "identifier"
    at org.adbcj.h2.H2DbException.<init>(H2DbException.java:22)
    at org.adbcj.h2.H2DbException.create(H2DbException.java:40)
    // SNIP

Simple Connection Pool

You can enable a connection pool by setting the org.adbcj.connectionpool.enable property to true.

When set to true, when you close a connection, it is returned to a connection pool. So, when connecting the next time, a initial connection handshakes can be skipped.

Map<String, String> settings = new HashMap<>();
settings.put(StandardProperties.CONNECTION_POOL_ENABLE, "false");
final ConnectionManager connectionManager = ConnectionManagerProvider.createConnectionManager(
        "adbcj:h2://localhost:14242/mem:db1;DB_CLOSE_DELAY=-1;MVCC=TRUE",
        "adbcj",
        "password1234",
        settings
);

Current implementation limitations:

  • At the moment there are no limits like timeouts, maximum connections.
  • If prepared statements or similar things were not closed, they are not closed when the connection returns to the pool. The prepared statements will leak on the database side.

Request Queue Length

With ADBCJ it is very easy to send hundreds to queries to the database, without waiting for results. You might queue up more queries than you expect and unexpectedly overload the database.

To protect from database overloads, ADBCJ as a default maximum queue length of 64 per connection That means if 64 requests are pending, and you issue another one, that one is aborted and a exception is returned.

An example:

// Blocking code to keep demo code short
Connection connection = connectionManager.connect().get();
PreparedQuery stmt = connection.prepareQuery("SELECT ?").get();

// Issue 200 requests, without waiting for any response
for (int i = 0; i < 200; i++) {
    stmt.execute(i).handle((rows, queryFailure) -> {
        if (queryFailure == null) {
            System.out.println(" - Yeah, Result: " + rows.get(0).get(0));
        } else {
            failures.incrementAndGet();
            System.err.println(" - Oups, we send to many queries");
            queryFailure.printStackTrace(System.err);
        }
        waitForCompletion.countDown();
        return null;
    });
}

This will almost certainly hit the que maximum, creating this error:

org.adbcj.DbException: To many pending requests. The current maximum is 64. Ensure that your not overloading the database with requests. Also check the adbcj.maxQueueLength property
    at org.adbcj.h2.H2Connection.failIfQueueFull(H2Connection.java:192)
    at org.adbcj.h2.H2Connection.queRequest(H2Connection.java:182)
    at org.adbcj.h2.H2PreparedQuery.executeWithCallback(H2PreparedQuery.java:29)
    at org.adbcj.PreparedQuery.execute(PreparedQuery.java:15)
    at info.adbcj.demo.TutorialQueueLimit.runQueries(TutorialQueueLimit.java:69)
    at info.adbcj.demo.TutorialQueueLimit.firstRunToSmallQueue(TutorialQueueLimit.java:39)
    at info.adbcj.demo.TutorialQueueLimit.main(TutorialQueueLimit.java:19)
Caused by: org.adbcj.DbException: To many pending requests. The current maximum is 64. Ensure that your not overloading the database with requests. Also check the adbcj.maxQueueLength property
    ... 7 more

You can increase the queue limit by setting the max queue length:

    int maxQueueLength = 200;
    Map<String, String> props = new HashMap<>();
    props.put(StandardProperties.MAX_QUEUE_LENGTH, String.valueOf(maxQueueLength));
    final ConnectionManager connectionManager = ConnectionManagerProvider.createConnectionManager(
            "adbcj:h2://localhost:14242/mem:db1;DB_CLOSE_DELAY=-1;MVCC=TRUE",
            "adbcj",
            "password1234",
            props
    );

Packages

No packages published

Languages

  • Java 99.9%
  • Shell 0.1%