I have a couple of questions after reaching out before about the use of NServiceBus in an environment that doesn’t require code to put a message on the servicebus.
First of all I was happy to hear that it is indeed possible to do this in our solution. Let me sketch you a bit of what we will want to do.
Our client is making use of another party for hosting the solution in Azure. They work with your product, mostly for using the nice logging and tracing capabilities that you have in the form of ServicePulse and ServiceInsight. They showed us how they worked with that and that scared us for a second, as we saw lots of code that felt that for the amount of different implementations that we will do, it would cause a lot of repetition.
What we want to use, is Logic Apps that get information from one system whenever another system requires it by letting the target system ask for the information. Basically the Logic App gets the request for information and puts a message on the ServiceBus in a topic that the source system is subscribed to. The source system replies with the information through another Logic App that puts the message back on the ServiceBus in a topic that the target system is subscribed to.
We will have a couple of those and because of this I have the following questions:
- How do we need to structure a message on the ServiceBus, in order for NServiceBus to understand that it has to pick it up? Message properties or headers that have to be set for this basically.
- How does Endpoint Configuration work for messages that have been sent directly to the ServiceBus? ServicePulse shows its monitoring based on Endpoints if I understand correctly.
- Can NServiceBus monitor several ServiceBus instances at the same time?
I hope my questions are clear, the matter is getting a little urgent as it is key to our implementation of a solution that is quite time sensitive.