On 24/01/2020 9:54 am, Francesco Raviglione wrote:
The "send.c" example is creating a proactor ("pn_proactor()"), then binding
it to a new connection using "pn_proactor_connect2()". My first question
is: is it possible to bind two connections to the same proactor? If yes,
how? Should I use somewhere the "pn_connection_wake()" function and use two
"run()" loops as shown in "send.c"?
You can execute the equivalent of that run() loop in two concurrent
threads. The proactor will ensure that events for the same connection
are not returned to both threads concurrently.
You only need to call pn_connection_wake() if you want to initiate some
action on that connection that is otherwise idle. (E.g. if there is now
something to send).
[...]
Moreover, I read that, in order to set the TTL for a message, I can use
"pn_message_set_ttl()". However, by compiling and running "send.c", I
noticed that, if a certain queue (specified, if I understood it correctly,
as "app.amqp_address"/"argv[3]") does not yet exist on the broker, it is
created.
When working with AMQP 0.9 and RabbitMQ, there was a way to set the
expiration for that queue, other than for single messages, in order to make
it temporary without directly interacting with the broker. As I am not very
expert on AMQP 1.0, do you know if there is any way to do the same in AMQP
1.0, using the C API?
Unlike the earlier pre-1.0 iterations, AMQP 1.0 does not define a
configuration schema for queues etc, or prescribe particular properties
or behaviours for queues outside the ability to interoperably transfer
messages.
In other words how you configure your broker will depend on which broker
you are using.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
