Skip to content

082.md

Latest commit

 

History

History
37 lines (24 loc) · 3.53 KB

082.md

File metadata and controls

37 lines (24 loc) · 3.53 KB
Raw

Ream Meeting #82 Recap

Overview

This week's call spanned ream dependency and memory maintenance, reth execution-layer integration nearing its metrics phase, data-availability work — a decoupled-DA checker and the proposer's data-column-sidecar broadcast, plus a Sepolia sync mismatch — and a cluster of EPF cohort-7 testing proposals covering Kurtosis-based beacon compliance, black-box consensus testing, and a spec-test oracle, several of which are being presented at tomorrow's office hours.

Development Progress

Ream Client Maintenance

  • Dependencies & memory: [Kayden] Updated ream's dependencies and spent time debugging a memory issue.

Execution-Layer Integration

  • reth integration: [Soham] Basic implementation is done; ran local devnets, found and fixed a few issues (now up for Sahil's review), and will start on metrics this week.

Data Availability

  • Decoupled DA: [Perfogic] Opened a decoupled-DA PR implementing a DA checker to handle the case where a parent block's data availability arrives after a child block's; broke down related issues in the ream repo as part of the EPF cohort-7 proposal.
  • Sepolia sync mismatch: [Perfogic] Syncing a ream node against Sepolia surfaced a mismatch — Sepolia is on Fulu, so the node tries to download the now-outdated blob sidecar instead of the data column sidecar; filed an issue that Hans is taking over. This week: focusing on the DA part of the ream beacon client.
  • Proposer data-column broadcast: [Hans] Finished the PR for the proposer to broadcast the data column sidecar on the beacon node, and is now root-causing the Sepolia sync / validate_attestation issue Perfogic reported — currently suspecting a configuration mismatch in how blob data is initialized on receipt.

Testing & Tooling

  • Kurtosis beacon compliance: [Tosin] Implemented the Prometheus metrics for beacon-client compliance and wrote a short doc for others doing similar work; fixing a few remaining beacon-client issues before the PR is ready to merge.
  • Black-box consensus testing: [Mohit] Incorporating Mario's review of the black-box testing proposal, prepared a mental model with pseudo-code for the planned test cases, and has three validation-suite test cases ready to PR later this week; presenting the project at tomorrow's office hours.
  • Spec-test oracle: [Richard] Finished and submitted the proposal (and presentation docs) to strengthen the spec-test oracle for the lean consensus, and got the go-ahead to present tomorrow; plans a backward-compatible state-transition schema using optional struct fields so each client can implement the endpoint (implementing the Rust-client endpoints during the fellowship). This week: sharing fixture-pinning findings with the fixture authors to settle whether the sparse pinnings are design choices or genuine gaps.

Next Steps

  • Kayden: Continue debugging the memory issue.
  • Soham: Address Sahil's review of the reth integration and start on integration metrics.
  • Perfogic: Focus on the DA part of the ream beacon client.
  • Hans: Find the root cause of the Sepolia sync / validate_attestation issue (blob-data initialization config mismatch).
  • Tosin: Resolve the remaining beacon-client issues and get the Prometheus-metrics PR ready to merge.
  • Mohit: Incorporate Mario's review, present at office hours, and open the PR for the three validation test cases.
  • Richard: Share the fixture-pinning findings with the fixture authors and coordinate the backward-compatible state-transition schema; present at office hours.