Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

While V2 -> V2 is also valuable, this is the more complex flow and therefore
the most valuable to test.

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

POST to /spi/reset will perform a soft restart of everything, including
reloading the config from disk.

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Ensure the multiparty contract is configured early enough for the
network version to be inspected during namespace init.

Initialize ff_system if ANY namespace uses multiparty V1 (not just
the default namespace).

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Allows callback handlers to wait on event processing to finish, if
desired.

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Do not attempt to stop the ff_system orchestrator - it's actually possible that
there may be unaggregated pins prior to the "terminate" event. Therefore, we
must continue listening indefinitely even if the handler is likely to be idle.

Also added some TODOs for next steps.

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Alex Shorsher <alex.shorsher@kaleido.io>

Signed-off-by: Alex Shorsher <alex.shorsher@kaleido.io>

Signed-off-by: Alex Shorsher <alex.shorsher@kaleido.io>

Signed-off-by: Alex Shorsher <alex.shorsher@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Also remove duplicate caching of network version.

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

…ates

If a subscription was created with a valid namespace name, and a V1 BatchPin
event is received on that subsscription with a different namespace name
embedded, simply drop it at the plugin layer. This will avoid passing an
extra de-duplication burden up to the event managers.

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Because ff_system is global, the contract instance is global. The address
can never change, and a second V1 contract can never be used. Adjust the
core logic to enforce this.

This also means splitting the E2E suite, as a V2->V2 migration is simple
to test over and over, but a V1->V2 migration can only be done once per
stack. This special test will have to be run infrequently under controlled
circumstances.

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

This is already handled by namespace manager, not orchestrator.

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>

Pass around a pointer to orchestrator and any children, but retain a single
copy of the namespace owned by namespace manager. Also use pointers for the
contract information to avoid unnecessary copies.

Signed-off-by: Andrew Richardson <andrew.richardson@kaleido.io>