Skip to content

test: MPC + stream component test (part 1)#757

Draft
jakmeier wants to merge 7 commits intosig-net:developfrom
jakmeier:stream_component_test_part1
Draft

test: MPC + stream component test (part 1)#757
jakmeier wants to merge 7 commits intosig-net:developfrom
jakmeier:stream_component_test_part1

Conversation

@jakmeier
Copy link
Copy Markdown
Contributor

Create a MockStream that connects to the MPC fixture
setup. This lets us test the indexer stream <-> MPC glue.

This is part 1, with just the tests adjusted to pass with the current configuration.

See #739 for tests that show current issues that need fixing.

@jakmeier jakmeier changed the title test: MPC + stream component test test: MPC + stream component test (part 1) Apr 16, 2026
@jakmeier jakmeier force-pushed the stream_component_test_part1 branch from b26e7fd to 2a87771 Compare April 23, 2026 18:57
@jakmeier jakmeier marked this pull request as ready for review April 30, 2026 15:06
@jakmeier jakmeier requested a review from volovyks April 30, 2026 15:08
@jakmeier
Copy link
Copy Markdown
Contributor Author

This is no longer blocked with #765 merged.

@volovyks I think you already reviewed most of this in the draft (#739).
This is the same but with test settings changed that they pass.

@jakmeier jakmeier marked this pull request as draft April 30, 2026 15:14
@jakmeier
Copy link
Copy Markdown
Contributor Author

@volovyks sorry, test_sign_no_presignature_waste apparently is still flaky in CI, I will ping you again when it's resolved

@jakmeier
Copy link
Copy Markdown
Contributor Author

just a quick update: it succeeds locally but fails in CI, probably it is the same problem as with test_channel_contention_multiple_blocks_at_once_delayed in #739. Round alignment problems. It worked in CI with ACCEPT_POSIT_TIMEOUT = 500ms but fails again with ACCEPT_POSIT_TIMEOUT = 100ms.

Basically, the current solution is still quite fragile. I need to look in to this more to decide if we can just increase the timeout for the tests, or if we can solve the underlying problem. (100ms for network delay should be plenty, even on slow CI machines)

jakmeier added 7 commits May 7, 2026 22:01
Create a MockStream that connects to the MPC fixture
setup. This lets us test the indexer stream <-> MPC glue.
but not with channel capacity?
otherwise it is always the same participant in round 0
adding useful cases and checking if they run in ci
@jakmeier jakmeier force-pushed the stream_component_test_part1 branch from 7a1f599 to 9271e9c Compare May 7, 2026 20:05
@jakmeier
Copy link
Copy Markdown
Contributor Author

jakmeier commented May 7, 2026

in theory, this should be passing tests now... but I did a bad rebase (clean in git but causes rust conflicts)

The trait ChainStream was changed on develop. I will have to fix that.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant