# EPF5 Week-15: Beacon metrics specs ## Week overview - [X] Continued working on metrics for Lighthouse, Grandine and Teku. - [X] Discussed metrics specs with the teams in the PeerDAS breakout Room. - [X] Created a draft PR into the beacon-metrics repo. - [X] Receiving comments on the PR from the teams. - [X] Built an initial dashboard for Teku locally and got unexpected results. This week was crucial for my project. On a PeerDAS Breakout Room, the DevOps team proposed the idea of metrics specs which my project stands for. Currently, the details are under discussion but the idea has been approved by the teams. Having specs improves and speeds up the testing phase. Clients can have a unified dashboard on an early testing stage while using Kurtosis. They can analyze it on the go and adjust the implementation of new EIPs. In the long-term perspective, It can shorten the whole testing phase and, as a result, fork delivery. ### Lighthouse I fetched the Lighthouse code updates for ssz and implemented req/resp metrics. Haven't pushed a PR yet as I'm currently making the final review of the code. ### Grandine Made code corrections after the review of [my PR](https://github.com/hangleang/grandine/pull/2). Worked with Hangleang closely. The PR has been merged. Now I can check the Grandine dashboard. ### Teku I implemented the first two PeerDAS metrics and pushed them into the [PR](https://github.com/Nashatyrev/teku/pull/139). The team wishes to make PRs small to check them easily. I built the Teku dashboard locally and got unexpected results - the negative values. It seems like an implementetion error. That's what we are going to investigate with the Teku team. ![Screenshot 2024-09-23 at 09.34.52](https://hackmd.io/_uploads/HJcrm3CaC.png) While implementing the Teku dashboard, I realized that having unified specs will let us use the same Kurtosis dashboard for all teams from the very beginning of implementing a new EIP. In my opinion, this approach can improve and shorten the testing phase and consequently the fork delivery in general. ## Weekly EPF Activity - [EPF5 Standup]() - [EPF5 Office Hours with Parithosh](https://github.com/eth-protocol-fellows/cohort-five/issues/436) - [PeerDAS Breakout Room]() - [ACDC #142](https://www.youtube.com/live/MRtJSnBU3Gk) [Github issue #1154](https://github.com/ethereum/pm/issues/1154) Devnet-3, Pectra discussion - split into PectrA and PectrB, PeerDAS/Blob scaling [Pectra scheduling into two forks by Alex Stokes](https://hackmd.io/@ralexstokes/rJVuKtlpR#Pectra) ## TODO - [ ] Build Grandine and Teku dashboards. - [ ] Compare the results of the three clients: Lighthouse, Teku an Grandine on existing metrics. - [ ] Continue working on metrics implementation. - [ ] Continue discussions on the specs with the teams. ## Links [Draft beacon metrics specs PR](https://github.com/ethereum/beacon-metrics/pull/13) [Grandine PeerDAS metrics PR](https://github.com/hangleang/grandine/pull/2) [Teku PeerDAS metrics PR](https://github.com/Nashatyrev/teku/pull/139)