Incorrect payload slot notifications #14
Labels
bug
Something isn't working
Comments
3 tasks
ferranbt
added a commit
to flashbots/rbuilder
that referenced
this issue
Aug 26, 2024
## 📝 Summary This PR does not fix the e2e test but makes it a bit more deterministic to execute. I feel the problem is that due to flashbots/builder-playground#14 issue, the `rbuilder` it not consistent in building blocks at the correct time for the playground. So, it misses building some blocks. The previous e2e test would send the transaction to the EL node (to be picked up by the builder) and the acceptance condition was for the transaction to be mined and that block to be mined by the local builder. However, since the builder was not consistent, the transaction would be mined by the local EL node instead and trigger the acceptance check. In this fix, I am only sending the transaction to the builder, so that the EL node cannot include it. Then, if the transaction gets mined, it was for sure done by the builder. I have tested it locally multiple times (>10) and it only failed once and I think due to flashbots/builder-playground#14. ## 💡 Motivation and Context <!--- (Optional) Why is this change required? What problem does it solve? Remove this section if not applicable. --> --- ## ✅ I have completed the following steps: * [x] Run `make lint` * [x] Run `make test` * [x] Added tests (if applicable)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
When the playground starts. Reth takes 2 slot iterations to sync up at startup. After that, the lighthouse sse stream does not return a consistent stream of event_payload messages every 12 seconds but a mix-up of different slot notifications.
This is not a problem for the
rbuilderbecause it will build blocks anyway but it is not a consistent behavior for the devnet.The text was updated successfully, but these errors were encountered: