# Week 9 ## PeerDAS Hi ,Rahul here this week I updated the [backports](https://mirror.xyz/stackrlabs.eth/5RuK1mWxVCuh6pIpPlqHLuofLAM0dW4N6ccbdvhNhOI) from the lighthouse crate of lighthouse to grandine's eth2libp2p and upgraded it to the latest version. I also got in touch with the other fellow on the progress made on the PeerDas implementation in grandine and the problem we need to solve is why the nodes are disconnecting. ## Light CLient The prysm client uses the bolt database for its operations , tho the original one has been archieved by the author so it uses a fork of the [bolt database](https://github.com/etcd-io/bbolt). I was working on adding db functions to store the LC updates. [This](https://www.youtube.com/watch?v=85MeiMA4dD8&t=610s) is an overral video on how serving LC endpoints would strain networking in the client. ![image](https://hackmd.io/_uploads/Bkrikhvc0.png) This is how much data needs to be served. We are planning to develop an app to automatically check the data from each client's light client endpoints to check for inconsistensies. # Plan for the next week - Work on the rest of the db functions in prysm - Continue to debug the grandine node disconnection problem