introducing an alternative to waitForBlock
#181
Comments
|
I suggest that the title should be something like: waitForBlock(N) is unreliable for synchronization As stated, any alternative would address the issue, including one that was functionally identical but refactored in some way. |
First implement it locally in a proposal test. If it works well then we can incorporate it upstream.
This would be ideal so the existing uses get automatically improved and no tests have to change. |
|
Thanks, I'm clear now. |
anilhelvaci
added a commit
to bytepitch/agoric-sdk
that referenced
this issue
Sep 30, 2024
mergify bot
added a commit
to Agoric/agoric-sdk
that referenced
this issue
Oct 22, 2024
refs: https://github.com/Agoric/BytePitchPartnerEng/issues/10 refs: Agoric/agoric-3-proposals#181 ## Description Currently the way to make sure a transaction sent to agoric chain has been executed is to use a method from `@agoric/synthetic-chain` called `waitForBlock` which stops the execution flow of a given ava test until N number of blocks are produced. This is not as deterministic as we'd like it to be but so far it worked fine. However, it could still end up in a race condition in some extreme cases. In this PR we introduce a set of tools to sync operations carried out in a test. Currently this PR supports operations like; * Making sure a core-eval resulted successfully deploying a contract * Making sure a core-eval successfully sent zoe invitations to committee members for governance * Making sure an account is successfully funded with vbank assets like IST, BLD etc. * Making sure an offer resulted successfully See Agoric/agoric-3-proposals#181 for further discussion. ### Security Considerations This is merely a testing tool, so no security considerations. ### Scaling Considerations This is merely a testing tool, so no scaling considerations. ### Documentation Considerations If this ends up getting in to `@agoric/synthetic-chain` then we might need more extensive explanation of what we do. But for now, I don't think any documentation is needed. ### Testing Considerations Have confirmed that this works fine with existing a3p tests locally but to keep the scope tighter I will update those tests one this PR lands. The coverage of the unit tests in `sync-tools.test.js` seems fine but I'm open to suggestions, of course.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Context
@dckc has raised in here that
waitForBlockmight be racy and pointed some good implementations for an alternative:Considerations
What do you think if we include one of the above to
@agoric/synthetic-chain? @turadgThe text was updated successfully, but these errors were encountered: