Introduction: You can find a description of this phase in section Phase 1 (Introduction).

(1) Only relevant for the add-on module Load Balancing to start the profile on a specific node. Note: See also section File Names, File Patterns, Paths, System Constants and Variables.

(2) Selection of an AMQP alias. See also section AMQP Connection.

(3) Selection of the message type (Subscriber, Topic, Routing or RPC). The style RPC is synchronous, the others are asynchronous. Note: If you use style RPC, it is additionally necessary to use a Response Route of type Custom Class and the specific class PassBackDataResponse. The reason is that in the case of an RPC call a response is expected by the calling client, which is generated by the mentioned Response Route class in Lobster_data and from there returned to the Input Agent and then to the client. See also section Third Possibility - HTTP Response Chain.

(4) Name of the queue or topic to listen to. See also (3). Note: If an Azure Service Bus is used and Topic is selected in (3), the following syntax must be used here: <topic_name>/subscriptions/<subscription_name>, e.g. mytopic/subscriptions/mysubscription.

(5) The routing key is a message attribute that the server uses to decide how the message should be routed to queues. For details please refer to the documentation of the AMQP server used, e.g. here for RabbitMQ. Note: Does not apply to type Subscriber and RPC.

(6) Set this checkbox if (4) is a durable queue (such a queue will survive a broker restart). See also section AMQP Connection.

(7) Here you can set the Consumer Delivery Acknowledgement behaviour, i.e. how Lobster_data confirms the receipt of a message to the message bus, see here, for example. If the checkbox is not set, the default behavior applies, i.e. the receipt is acknowledged after the backup file of the profile job has been created (except for message type RPC). If the checkbox is set, a received message will only be confirmed after the job has been successfully completed. Processing in the background is not possible in that case (parallel processing and queueing of jobs). If the Response Routes have checkbox Execute response in own thread set, the receipt of a message is always confirmed after phase 6. Note: If a message is not confirmed by Lobster_data, the setting of the AMQP server takes effect and the message may be delivered again or rejected and forwarded to the Dead Letter Queue (see also the following section Message Properties). If this is the case, you can therefore, similar to working with databases, "commit" a transaction at the end of the processing in the profile if no error has occurred. In the event of an error, profiles for error handling can be set up, for example, which may then process the erroneous messages from the dead letter queue. Note: A dead letter queue can also be created via an AMQP Response Route if the profile that uses this Response Route has a defined system variable AMQP_SYS_<name of the message system property>. For RabbitMQ, for example, that would be AMQP_SYS_x-dead-letter-exchange, see here.

(8) Here you can enter the name of a message property. The value of the property is then used as the name of the received file (log overview in the Control Centre). If no value is specified or the property is not found or has no value, the default value amqp.data is used for the file name. Example: The received message has the message property myfilename with the value test.txt. If the field has the value myfilename, the received file gets the name test.txt. Note: Please note that you cannot use the file name directly in the field.

(9) and (10) Messages are collected in a buffer. The profile is started with the collected messages when either the maximum number of collected messages is reached or the maximum waiting time. If the profile is saved, messages in the buffer are deleted. Note: The desired seconds can also be entered manually.

(11) If messages are collected (value in (9) is greater than 1), you can decide which ones are to be passed on. If you select All, you might have to use a delimiter (12) between the single messages.

(12) The delimiter for messages. You can also specify hex codes with prefix 0x. See also (11). Note: Only active if the value in (9) is greater than 1.

(13) Here you can specify conditions (for numerical values, for strings or regular expressions) when messages are to be forwarded to the profile for processing prematurely (only one must apply). If, in the example, the message does not contain the string lens, it will be collected. If the string lens is contained, the message is passed on immediately and the collection counter is reset, i.e. other collected messages are discarded. As an alternative to conditions, you can use a function chain to check messages. If the function chain returns true at the end, the check is considered successful. Note: Only active if the value in (9) is greater than 1.

(14) Enter the number of messages here until the function chain is checked. 1 if every message should be checked.

(15) Select a function by double-clicking (or using the arrow). It will be inserted in (16).

(16) The function chain, as you already know it from phase 3. Note: Note the specific parameter type Message.

(17) In addition to the parameter types already known from phase 3, there will also be the type Message.

(18) You can test your function chain. See the following screenshot. Note: Either the calculated value or a fixed value can be used as the result of each individual function.

(19) Here you can manually enter a message that will be used for the test.

(20) Click here to run the test.

(21) The result of the test.