Add E2E tests for contract termination

[awrichar]

Part of FIR-12

Depends on hyperledger/firefly-cli#197
In a chain with #904

Add a new E2E test which dynamically:

• deploys two instances of the FireFly multiparty contract
• dynamically adds a new namespace to the config of both nodes
• resets both nodes to read the config
• sends one message on the first contract
• performs a network migration
• sends one message on the new contract

There are variations of this test for both V1->V2 contract migrations and V2->V2. Notably, the V1->V2 migration can only be performed once per stack, so it is not part of the main suites. We may need to add a dedicated nightly job for this test.

This PR also includes loads of fixes for things exposed by the E2E tests, including major changes to how the legacy ff_system namespace is managed. There can now only be one multiparty V1 contract EVER used in the history of a FireFly node, and attempting to configure any namespace with a different V1 contract is an error. The ff_system orchestrator will be initialized with that V1 config (if any is found) and will live forever (walking back some of the changes from #884), because we can never be fully sure when all pins from ff_system have been aggregated.

[codecov-commenter]

Codecov Report

Merging #899 (33bf296) into main (ce9d5a5) will decrease coverage by 0.01%.
The diff coverage is 100.00%.

❗ Current head 33bf296 differs from pull request most recent head f1c6bc9. Consider uploading reports for the commit f1c6bc9 to get more accurate results

@@            Coverage Diff             @@
##             main     #899      +/-   ##
==========================================
- Coverage   99.98%   99.97%   -0.02%     
==========================================
  Files         297      298       +1     
  Lines       19400    19463      +63     
==========================================
+ Hits        19397    19458      +61     
- Misses          3        4       +1     
- Partials        0        1       +1     

Impacted Files	Coverage Δ
internal/orchestrator/bound_callbacks.go	100.00% <ø> (ø)
pkg/core/namespace.go	100.00% <ø> (ø)
cmd/firefly.go	100.00% <100.00%> (ø)
internal/apiserver/route_get_namespace.go	100.00% <100.00%> (ø)
internal/apiserver/route_spi_post_reset.go	100.00% <100.00%> (ø)
internal/blockchain/ethereum/ethereum.go	100.00% <100.00%> (ø)
internal/blockchain/fabric/fabric.go	100.00% <100.00%> (ø)
internal/events/network_action.go	100.00% <100.00%> (ø)
internal/multiparty/manager.go	100.00% <100.00%> (ø)
internal/namespace/manager.go	100.00% <100.00%> (ø)
... and 4 more

Help us with your feedback. Take ten seconds to tell us how you rate us.

[awrichar]

Factored some of the general enhancements out to #904 as a pre-req before adding the new test here.

[awrichar]

Further, honing of this test has revealed that the migration feature is still a bit broken. Moving back to draft state while addressing the remaining issues.

[awrichar]

Fixed up the test so that it passes now, so I'm opening this back up for review. I think there may still be work to do to reliably terminate the ff_system namespace though.

[shorsher]

Trying to fix the merge conflicts + update tests got a bit gnarly. Going to fix the tests in this branch, make the TODO changes, and then fix the conflicts.

[shorsher]

tests updated, pulling main in now.

[shorsher]

Currently investigating the contract migration test and why firefly does not seem to pick up new namespaces after a reset.

[awrichar]

This is now ready for review again. Concluded that the ff_system namespace actually can never be "stopped". Once it exists, we have to keep the listener around forever.

[shorsher]

🚀