wip: upgrade statemine/t integration tests to xcm v3 #2224
Conversation
but not quite for the right reasons...
EgorPopelyaev
added
B0-silent
| @@ -170,245 +77,76 @@ tests: | |||
| - name: Penpal Parachain sends a request to the Relay Chain to open a channel with the Assets Parachain | |||
There was a problem hiding this comment.
Seems like the context has changed here. We are no longer sending the request from Penpal, but rather we call an extrinsic on the relay chain to open the channel. I guess we should then change the name of this test to something that reflects this scenario better.
There was a problem hiding this comment.
Got it fixed in parachains-integration-tests repo.
| chain: *relay_chain | ||
| pallet: hrmp | ||
| call: hrmpOpenChannelRequestsList | ||
| args: [] | ||
| - asserts: | ||
| equal: | ||
| args: [ | ||
| $requested_channels, | ||
| $requested_channels_a_to_p, |
There was a problem hiding this comment.
This assertion always fails for me. And it totally depends on the sequence of execution, if we change those tests around - it still fails. Because the list is empty. Since request processing happens on session start, this is going to be flaky even for the first test, depending on the machine and other variables. I suggest calling force_process_hrmp_open and then querying HrmpIngressChannelsIndex or HrmpEgressChannelsIndex. Otherwise we could just remove the assert completely as the receipt of HrmpChannelForceOpened has to be enough.
There was a problem hiding this comment.
Agreed I think we don't need that assertion - as long as we have the event I think we're good. The tests in the next file would fail if the channel was not open.
|
How will you test new kusama runtime works with both old statemine runtime and new statemine runtime? |
|
Closing this PR as most of it is put into paritytech/parachains-integration-tests#92 - we will assume that is master for the tests from now on. |
(best to look commit by commit to reduce from initially updating the tests from the other repo)