Handle failures when starting node on integration tests

[mFragaBA]

What should be done?

When the cargo make start-node fails we have no way of detecting that and ending the test prematurely. This makes the test run until it time outs (and because of nextest it'll run 3 times like that).

It would also be nice to be able to read the logs from the node tasks.

How should it be done?

Currently the best approach I could find was to split the start-node into 2:

• first do everything up to the miden-node-command ... start
• lastly, do call miden-node-command ... start &

That should handle most breaking issues. We have to run the node on background so the workflow continues with the next step.

A secondary issue is that we can't see the node's output. One way to handle that is by using github's artifacts. We can either preserve each log file or output it in a later workflow step after executing tests (not sure what happens if the integration test fails though)

When is this task done?

when we can detect failures when starting the node and ending the test prematurely

Additional context

No response

[igamigo]

As part of this, we should also bring our own node configuration files instead of relying on the node-provided ones. I don't think we use accounts generated on genesis on integration tests, for example.

[mFragaBA]

closing as #310 got merged