Chapter 5. Using the API

This chapter explains how to use the AMQ Python API to perform common messaging tasks.

For more information, see the AMQ Python API reference and AMQ Python example suite.

5.1. Basic operation

5.1.1. Handling messaging events

AMQ Python is an asynchronous event-driven API. To define how an application handles events, the user implements callback methods on the MessagingHandler class. These methods are then called as network activity or timers trigger new events.

Example: Handling messaging events

class ExampleHandler(MessagingHandler):
    def on_start(self, event):
        print("The container event loop has started")

    def on_sendable(self, event):
        print("A message can be sent")

    def on_message(self, event):
        print("A message is received")

These are only a few common-case events. The full set is documented in the API reference.

The event argument has attributes for accessing the object the event is regarding. Attributes with no relevance to a particular event are null.

Example: Accessing event objects

event.container
event.connection
event.session
event.sender
event.receiver
event.delivery
event.message

5.1.2. Creating a container

The container is the top-level API object. It is the entry point for creating connections, and it is responsible for running the main event loop. It is often constructed with a global event handler.

Example: Creating a container

handler = ExampleHandler()
container = Container(handler)
container.run()

Setting the container identity

Each container instance has a unique identity called the container ID. When AMQ Python makes a connection, it sends the container ID to the remote peer. To set the container ID, pass it to the Container constructor.

Example: Setting the container identity

container = Container(handler)
container.container_id = "job-processor-3"

If the user does not set the ID, the library will generate a UUID when the container is constucted.

5.2. Network connections

5.2.1. Connection URLs

Connection URLs encode the information used to establish new connections.

Connection URL syntax

scheme://host[:port]

  • Scheme - The connection transport, either amqp for unencrypted TCP or amqps for TCP with SSL/TLS encryption.
  • Host - The remote network host. The value can be a hostname or a numeric IP address. IPv6 addresses must be enclosed in square brackets.
  • Port - The remote network port. This value is optional. The default value is 5672 for the amqp scheme and 5671 for the amqps scheme.

Connection URL examples

amqps://example.com
amqps://example.net:56720
amqp://127.0.0.1
amqp://[::1]:2000

5.2.2. Creating outgoing connections

To connect to a remote server, call the Container.connect() method with a connection URL. This is typically done inside the MessagingHandler.on_start() method.

Example: Creating outgoing connections

class ExampleHandler(MessagingHandler):
    def on_start(self, event):
        event.container.connect("amqp://example.com")

    def on_connection_opened(self, event):
        print("Connection", event.connection, "is open")

See the Section 5.5, “Security” section for information about creating secure connections.

5.2.3. Listening for incoming connections

AMQ Python can accept inbound network connections, allowing you to build messaging servers. To start listening for connections, use the Container.listen() method with a URL containing the local host address and port to listen on.

Example: Listening for incoming connections

class ExampleHandler(MessagingHandler):
    def on_start(self, event):
        event.container.listen("0.0.0.0")

    def on_connection_opened(self, event):
        print("New incoming connection", event.connection)

The special IP address 0.0.0.0 listens on all available IPv4 interfaces. To listen on all IPv6 interfaces, use [::0].

For more information, see the server receive.py example.

5.2.4. Configuring reconnect

Reconnect allows a client to recover from lost connections. It is used to ensure that the components in a distributed system reestablish communication after temporary network or component failures.

AMQ Python enables reconnect by default. If a connection is lost or a connection attempt fails, the client will try again after a brief delay. The delay increases exponentially for each new attempt, up to a default maximum of 10 seconds.

To disable reconnect, set the reconnect connection option to False.

Example: Disabling reconnect

container.connect("amqp://example.com", reconnect=False)

To control the delays between connection attempts, define a class implementing the reset() and next() methods and set the reconnect connection option to an instance of that class.

Example: Configuring reconnect

class ExampleReconnect(object):
    def __init__(self):
        self.delay = 0

    def reset(self):
        self.delay = 0

    def next(self):
        if self.delay == 0:
            self.delay = 0.1
        else:
            self.delay = min(10, 2 * self.delay)

        return self.delay

container.connect("amqp://example.com", reconnect=ExampleReconnect())

The next method returns the next delay in seconds. The reset method is called once before the reconnect process begins.

5.2.5. Configuring failover

AMQ Python allows you to configure multiple connection endpoints. If connecting to one fails, the client attempts to connect to the next in the list. If the list is exhausted, the process starts over.

To specify multiple connection endpoints, set the urls connection option to a list of connection URLs.

Example: Configuring failover

urls = ["amqp://alpha.example.com", "amqp://beta.example.com"]
container.connect(urls=urls)

It is an error to use the url and urls options at the same time.

5.3. Senders and receivers

The client uses sender and receiver links to represent channels for delivering messages. Senders and receivers are unidirectional, with a source end for the message origin, and a target end for the message destination.

Source and targets often point to queues or topics on a message broker. Sources are also used to represent subscriptions.

5.3.1. Creating queues and topics on demand

Some message servers support on-demand creation of queues and topics. When a sender or receiver is attached, the server uses the sender target address or the receiver source address to create a queue or topic with a name matching the address.

The message server typically defaults to creating either a queue (for one-to-one message delivery) or a topic (for one-to-many message delivery). The client can indicate which it prefers by setting the queue or topic capability on the source or target.

To select queue or topic semantics, follow these steps:

  1. Configure your message server for automatic creation of queues and topics. This is often the default configuration.
  2. Set either the queue or topic capability on your sender target or receiver source, as in the examples below.

Example: Sending to a queue created on demand

class CapabilityOptions(SenderOption):
    def apply(self, sender):
        sender.target.capabilities.put_object(symbol("queue"))

class ExampleHandler(MessagingHandler):
    def on_start(self, event):
        conn = event.container.connect("amqp://example.com")
        event.container.create_sender(conn, "jobs", options=CapabilityOptions())

Example: Receiving from a topic created on demand

class CapabilityOptions(ReceiverOption):
    def apply(self, receiver):
        receiver.source.capabilities.put_object(symbol("topic"))

class ExampleHandler(MessagingHandler):
    def on_start(self, event):
        conn = event.container.connect("amqp://example.com")
        event.container.create_receiver(conn, "notifications", options=CapabilityOptions())

For more information, see the following examples:

5.3.2. Creating durable subscriptions

A durable subscription is a piece of state on the remote server representing a message receiver. Ordinarily, message receivers are discarded when a client closes. However, because durable subscriptions are persistent, clients can detach from them and then re-attach later. Any messages received while detached are available when the client re-attaches.

Durable subscriptions are uniquely identified by combining the client container ID and receiver name to form a subscription ID. These must have stable values so that the subscription can be recovered.

To create a durable subscription, follow these steps:

  1. Set the connection container ID to a stable value, such as client-1:

    container = Container(handler)
    container.container_id = "client-1"
  2. Configure the receiver source for durability by setting the durability and expiry_policy properties:

    class SubscriptionOptions(ReceiverOption):
        def apply(self, receiver):
            receiver.source.durability = Terminus.DELIVERIES
            receiver.source.expiry_policy = Terminus.EXPIRE_NEVER
  3. Create a receiver with a stable name, such as sub-1, and apply the source properties:

    event.container.create_receiver(conn, "notifications",
                                    name="sub-1",
                                    options=SubscriptionOptions())

To detach from a subscription, use the Receiver.detach() method. To terminate the subscription, use the Receiver.close() method.

For more information, see the durable-subscribe.py example.

5.3.3. Creating shared subscriptions

A shared subscription is a piece of state on the remote server representing one or more message receivers. Because it is shared, multiple clients can consume from the same stream of messages.

The client configures a shared subscription by setting the shared capability on the receiver source.

Shared subscriptions are uniquely identified by combining the client container ID and receiver name to form a subscription ID. These must have stable values so that multiple client processes can locate the same subscription. If the global capability is set in addition to shared, the receiver name alone is used to identify the subscription.

To create a durable subscription, follow these steps:

  1. Set the connection container ID to a stable value, such as client-1:

    container = Container(handler)
    container.container_id = "client-1"
  2. Configure the receiver source for sharing by setting the shared capability:

    class SubscriptionOptions(ReceiverOption):
        def apply(self, receiver):
            receiver.source.capabilities.put_object(symbol("shared"))
  3. Create a receiver with a stable name, such as sub-1, and apply the source properties:

    event.container.create_receiver(conn, "notifications",
                                    name="sub-1",
                                    options=SubscriptionOptions())

To detach from a subscription, use the Receiver.detach() method. To terminate the subscription, use the Receiver.close() method.

For more information, see the shared-subscribe.py example.

5.4. Message delivery

5.4.1. Sending messages

To send a message, override the on_sendable event handler and call the Sender.send() method. The sendable event fires when the Sender has enough credit to send at least one message.

Example: Sending messages

class ExampleHandler(MessagingHandler):
    def on_start(self, event):
        conn = event.container.connect("amqp://example.com")
        sender = event.container.create_sender(conn, "jobs")

    def on_sendable(self, event):
        message = Message("job-content")
        event.sender.send(message)

For more information, see the send.py example.

5.4.2. Tracking sent messages

When a message is sent, the sender can keep a reference to the delivery object representing the transfer. After the message is delivered, the receiver accepts or rejects it. The sender is notified of the outcome for each delivery.

To monitor the outcome of a sent message, override the on_accepted and on_rejected event handlers and map the delivery state update to the delivery returned from send().

Example: Tracking sent messages

def on_sendable(self, event):
    message = Message(self.message_body)
    delivery = event.sender.send(message)

def on_accepted(self, event):
    print("Delivery", event.delivery, "is accepted")

def on_rejected(self, event):
    print("Delivery", event.delivery, "is rejected")

5.4.3. Receiving messages

To receive a message, create a receiver and override the on_message event handler.

Example: Receiving messages

class ExampleHandler(MessagingHandler):
    def on_start(self, event):
        conn = event.container.connect("amqp://example.com")
        receiver = event.container.create_receiver(conn, "jobs")

    def on_message(self, event):
        print("Received message", event.message, "from", event.receiver)

For more information, see the receive.py example.

5.4.4. Acknowledging received messages

To explicitly accept or reject a delivery, use the Delivery.update() method with the ACCEPTED or REJECTED state in the on_message event handler.

Example: Acknowledging received messages

def on_message(self, event):
    try:
        process_message(event.message)
        event.delivery.update(ACCEPTED)
    except:
        event.delivery.update(REJECTED)

By default, if you do not explicity acknowledge a delivery, then the library accepts it after on_message returns. To disable this behavior, set the auto_accept receiver option to false.

5.5. Security

5.5.1. Securing connections with SSL/TLS

AMQ Python uses SSL/TLS to encrypt communication between clients and servers.

To connect to a remote server with SSL/TLS, use a connection URL with the amqps scheme.

Example: Enabling SSL/TLS

container.connect("amqps://example.com")

5.5.2. Connecting with a user and password

AMQ Python can authenticate connections with a user and password.

To specify the credentials used for authentication, set the user and password options on the connect() method.

Example: Connecting with a user and password

container.connect("amqps://example.com", user="alice", password="secret")

5.5.3. Configuring SASL authentication

AMQ Python uses the SASL protocol to perform authentication. SASL can use a number of different authentication mechanisms. When two network peers connect, they exchange their allowed mechanisms, and the strongest mechanism allowed by both is selected.

Note

The client uses Cyrus SASL to perform authentication. Cyrus SASL uses plug-ins to support specific SASL mechanisms. Before you can use a particular SASL mechanism, the relevant plug-in must be installed. For example, you need the cyrus-sasl-plain plug-in in order to use SASL PLAIN authentication.

To see a list of Cyrus SASL plug-ins in Red Hat Enterprise Linux, use the yum search cyrus-sasl command. To install a Cyrus SASL plug-in, use the yum install PLUG-IN command.

By default, AMQ Python allows all of the mechanisms supported by the local SASL library configuration. To restrict the allowed mechanisms and thereby control what mechanisms can be negotiated, use the allowed_mechs connection option. It takes a string containing a space-separated list of mechanism names.

Example: Configuring SASL authentication

container.connect("amqps://example.com", allowed_mechs="ANONYMOUS")

This example forces the connection to authenticate using the ANONYMOUS mechanism even if the server we connect to offers other options. Valid mechanisms include ANONYMOUS, PLAIN, SCRAM-SHA-256, SCRAM-SHA-1, GSSAPI, and EXTERNAL.

AMQ Python enables SASL by default. To disable it, set the sasl_enabled connection option to false.

Example: Disabling SASL

event.container.connect("amqps://example.com", sasl_enabled=False)

5.5.4. Authenticating using Kerberos

Kerberos is a network protocol for centrally managed authentication based on the exchange of encrypted tickets. See Using Kerberos for more information.

  1. Configure Kerberos in your operating system. See Configuring Kerberos to set up Kerberos on Red Hat Enterprise Linux.
  2. Enable the GSSAPI SASL mechanism in your client application.

    container.connect("amqps://example.com", allowed_mechs="GSSAPI")
  3. Use the kinit command to authenticate your user credentials and store the resulting Kerberos ticket.

    $ kinit <user>@<realm>
  4. Run the client program.

5.6. Logging

5.6.1. Enabling protocol logging

The client can log AMQP protocol frames to the console. This data is often critical when diagnosing problems.

To enable protocol logging, set the PN_TRACE_FRM environment variable to 1:

Example: Enabling protocol logging

$ export PN_TRACE_FRM=1
$ <your-client-program>

To disable protocol logging, unset the PN_TRACE_FRM environment variable.

5.7. Tracing

The client offers distributed tracing based on the Jaeger implementation of the OpenTracing standard. Use the following steps to enable tracing in your application:

  1. Install the tracing dependencies.

    Red Hat Enterprise Linux 7

    $ sudo yum install https://dl.fedoraproject.org/pub/epel/epel-release-latest-7.noarch.rpm
    $ sudo yum install python2-pip
    $ pip install --user --upgrade setuptools
    $ pip install --user opentracing jaeger-client

    Red Hat Enterprise Linux 8

    $ sudo dnf install python3-pip
    $ pip3 install --user opentracing jaeger-client

  2. Register the global tracer in your program.

    Example: Global tracer configuration

    from proton.tracing import init_tracer
    
    tracer = init_tracer("<service-name>")

For more information about Jaeger configuration, see Jaeger Sampling.

When testing or debugging, you may want to force Jaeger to trace a particular operation. See the Jaeger Python client documentation for more information.

To view the traces your application captures, use the Jaeger Getting Started to run the Jaeger infrastructure and console.