activemq-artemis/docs/user-manual/large-messages.adoc

271 lines
9.6 KiB
Plaintext

= Large Messages
:idprefix:
:idseparator: -
Apache ActiveMQ Artemis can be configured to give special treatment to messages which are beyond a configured size.
Instead of keeping the entire contents of these messages _in memory_ the broker will hold just a thin object on the queues with a reference to the content (e.g. in a file or a database table).
This is supported on Core Protocol and on the AMQP Protocol.
== Configuring the server
When using the xref:persistence.adoc#file-journal-default[file journal] large messages are stored on disk on the server.
The configuration property `large-messages-directory` specifies where large messages are stored.
[,xml]
----
<configuration...>
<core...>
...
<large-messages-directory>data/large-messages</large-messages-directory>
...
</core>
</configuration>
----
By default `large-messages-directory` is `data/largemessages`.
[NOTE]
====
For the best performance we recommend using the file journal with the large messages directory on a different physical volume to the message journal or paging directory.
====
For xref:persistence.adoc#jdbc-persistence[JDBC persistence] the `large-message-table` should be configured.
[,xml]
----
<configuration...>
<core...>
...
<store>
<database-store>
...
<large-message-table-name>LARGE_MESSAGES_TABLE</large-message-table-name>
...
</database-store>
</store>
...
</core>
</configuration>
----
By default `large-message-table` is `LARGE_MESSAGE_TABLE`.
By default when writing the final bytes to a large message all writes are synchronized to the storage medium.
This can be configured via `large-message-sync`, e.g.:
[,xml]
----
<configuration...>
<core...>
...
<large-message-sync>true</large-message-sync>
...
</core>
</configuration>
----
By default `large-message-sync` is `true`.
== Configuring the Core Client
Any message larger than a certain size is considered a large message.
Large messages will be split up and sent in fragments.
This is determined by the URL parameter `minLargeMessageSize`
[NOTE]
====
Apache ActiveMQ Artemis messages are encoded using 2 bytes per character so if the message data is filled with ASCII characters (which are 1 byte) the size of the resulting Apache ActiveMQ Artemis message would roughly double.
This is important when calculating the size of a "large" message as it may appear to be less than the `minLargeMessageSize` before it is sent, but it then turns into a "large" message once it is encoded.
====
The default value is 100KiB.
xref:configuring-transports.adoc#configuring-the-transport-directly-from-the-client[Configuring the transport directly from the client side] will provide more information on how to instantiate the core session factory or JMS connection factory.
== Compressed Large Messages on Core Protocol
You can choose to send large messages in compressed form using `compressLargeMessage` URL parameter.
If you specify the boolean URL parameter `compressLargeMessage` as true, the system will use the ZIP algorithm to compress the message body as the message is transferred to the server's side.
Notice that there's no special treatment at the server's side, all the compressing and uncompressing is done at the client.
This behavior can be tuned further by setting an optional parameter: `compressionLevel`.
This will decide how much the message body should be compressed.
`compressionLevel` accepts an integer of `-1` or a value between `0-9`.
The default value is `-1` which corresponds to around level 6-7.
If the compressed size of a large message is below `minLargeMessageSize`, it is sent to server as regular messages.
This means that the message won't be written into the server's large-message data directory, thus reducing the disk I/O.
NOTE: A higher `compressionLevel` means the message body will get further compressed, but this is at the cost of speed and computational overhead.
Make sure to tune this value according to its specific use-case.
== Streaming large messages from Core Protocol
Apache ActiveMQ Artemis supports setting the body of messages using input and output streams (`java.lang.io`)
These streams are then used directly for sending (input streams) and receiving (output streams) messages.
When receiving messages there are 2 ways to deal with the output stream;
you may choose to block while the output stream is recovered using the method `ClientMessage.saveOutputStream` or alternatively using the method `ClientMessage.setOutputstream` which will asynchronously write the message to the stream.
If you choose the latter the consumer must be kept alive until the message has been fully received.
You can use any kind of stream you like.
The most common use case is to send files stored in your disk, but you could also send things like JDBC Blobs, `SocketInputStream`, things you recovered from `HTTPRequests` etc.
Anything as long as it implements `java.io.InputStream` for sending messages or `java.io.OutputStream` for receiving them.
=== Streaming over Core API
The following table shows a list of methods available at `ClientMessage` which are also available through JMS by the use of object properties.
|===
| Name | Description | JMS Equivalent
| setBodyInputStream(InputStream)
| Set the InputStream used to read a message body when sending it.
| JMS_AMQ_InputStream
| setOutputStream(OutputStream)
| Set the OutputStream that will receive the body of a message.
This method does not block.
| JMS_AMQ_OutputStream
| saveOutputStream(OutputStream)
| Save the body of the message to the `OutputStream`.
It will block until the entire content is transferred to the `OutputStream`.
| JMS_AMQ_SaveStream
|===
To set the output stream when receiving a core message:
[,java]
----
ClientMessage msg = consumer.receive(...);
// This will block here until the stream was transferred
msg.saveOutputStream(someOutputStream);
ClientMessage msg2 = consumer.receive(...);
// This will not wait the transfer to finish
msg2.setOutputStream(someOtherOutputStream);
----
Set the input stream when sending a core message:
[,java]
----
ClientMessage msg = session.createMessage();
msg.setInputStream(dataInputStream);
----
Notice also that for messages with more than 2GiB the getBodySize() will return invalid values since this is an integer (which is also exposed to the JMS API).
On those cases you can use the message property _AMQ_LARGE_SIZE.
=== Streaming over JMS
When using JMS, Apache ActiveMQ Artemis maps the streaming methods on the core API (see ClientMessage API table above) by setting object properties . You can use the method `Message.setObjectProperty` to set the input and output streams.
The `InputStream` can be defined through the JMS Object Property JMS_AMQ_InputStream on messages being sent:
[,java]
----
BytesMessage message = session.createBytesMessage();
FileInputStream fileInputStream = new FileInputStream(fileInput);
BufferedInputStream bufferedInput = new BufferedInputStream(fileInputStream);
message.setObjectProperty("JMS_AMQ_InputStream", bufferedInput);
someProducer.send(message);
----
The `OutputStream` can be set through the JMS Object Property JMS_AMQ_SaveStream on messages being received in a blocking way.
[,java]
----
BytesMessage messageReceived = (BytesMessage)messageConsumer.receive(120000);
File outputFile = new File("huge_message_received.dat");
FileOutputStream fileOutputStream = new FileOutputStream(outputFile);
BufferedOutputStream bufferedOutput = new BufferedOutputStream(fileOutputStream);
// This will block until the entire content is saved on disk
messageReceived.setObjectProperty("JMS_AMQ_SaveStream", bufferedOutput);
----
Setting the `OutputStream` could also be done in a non blocking way using the property JMS_AMQ_OutputStream.
[,java]
----
// This won't wait the stream to finish. You need to keep the consumer active.
messageReceived.setObjectProperty("JMS_AMQ_OutputStream", bufferedOutput);
----
[NOTE]
====
When using JMS, Streaming large messages are only supported on `StreamMessage` and `BytesMessage`.
====
=== Streaming Alternative on Core Protocol
If you choose not to use the `InputStream` or `OutputStream` capability of Apache ActiveMQ Artemis You could still access the data directly in an alternative fashion.
On the Core API just get the bytes of the body as you normally would.
[,java]
----
ClientMessage msg = consumer.receive();
byte[] bytes = new byte[1024];
for (int i = 0 ; i < msg.getBodySize(); i += bytes.length)
{
msg.getBody().readBytes(bytes);
// Whatever you want to do with the bytes
}
----
If using JMS API, `BytesMessage` and `StreamMessage` also supports it transparently.
[,java]
----
BytesMessage rm = (BytesMessage)cons.receive(10000);
byte data[] = new byte[1024];
for (int i = 0; i < rm.getBodyLength(); i += 1024)
{
int numberOfBytes = rm.readBytes(data);
// Do whatever you want with the data
}
----
== Configuring AMQP Acceptor
You can configure the property `amqpMinLargeMessageSize` at the acceptor.
The default value is 102400 (100KBytes).
Setting it to -1 will disable large message support.
WARNING: setting amqpMinLargeMessageSize to -1, your AMQP message might be stored as a Core Large Message if the size of the message does not fit into the journal.
This is the former semantic of the broker and it is kept this way for compatibility reasons.
[,xml]
----
<acceptors>
<!-- AMQP Acceptor. Listens on default AMQP port for AMQP traffic.-->
<acceptor name="amqp">tcp://0.0.0.0:5672?; ..... amqpMinLargeMessageSize=102400; .... </acceptor>
</acceptors>
----
== Large message example
Please see the xref:examples.adoc#large-message[Large Message Example] which shows how large messages are configured and used with JMS.