Checkpoint sync for Eth testnets

[jgresham]

Feature details

• Use checkpoint sync for Sepolia and Holesky for Consensus Clients that already use it for mainnet
• Test it because some Consensus clients don't work with certain URLs

Checkpoint sync urls by network

[arionrefat]

@jgresham Hi, I want to work on it. Can you more pointers on how can I solve this ?

[jgresham]

@jgresham Hi, I want to work on it. Can you more pointers on how can I solve this ?

great! The work is mostly done in another repo, which serves each node config controls to NiceNode, and might just be setting the urls in the json files. You can run this api locally and have NiceNode use the local server to test. The format of these json files isn't really documented well however. A couple options:

1. set the checkpointSyncUrl: <url> in the Ethereum json file under

Sepolia config...
    "serviceId": "consensusClient", 
    "configValues": {
        "checkpointSyncUrl: <sepolia-checkpoint-sync>

https://github.com/NiceNode/controller-api/blob/main/src/app/api/controller/NodeSpecs/ethereum/ethereum-v1.0.0.json#L118

This should set the checkpointSyncUrl for each consensus client.
Also adding "addNodeFlow": "optional" to the checkpointSyncUrl: { } config for each consensus client should allow the user to see this when adding the node and before starting the node for the first time (important for checkpoint sync).

2. we could just add on the checkpoint sync flag for each network config like here https://github.com/NiceNode/controller-api/blob/main/src/app/api/controller/NodeSpecs/lodestar/lodestar-v1.0.0.json#L67

[jgresham]

Sometimes the checkpoint sync servers are slow or the local client doesn't wait long enough. Adjusting the timeout config for the client might be necessary (lighthouse docs https://lighthouse-book.sigmaprime.io/checkpoint-sync.html#adjusting-the-timeout)

[arionrefat]

@jgresham Thank you so much man. I will look into it and ask questions if require.

[arionrefat]

we could just add on the checkpoint sync flag for each network config like here https://github.com/NiceNode/controller-api/blob/main/src/app/api/controller/NodeSpecs/lodestar/lodestar-v1.0.0.json#L67

I went with option 2 and tested it out in nice-node

it seems it's synchronising for Sepolia and Holesky for Consensus Clients.

Am I doing it right? @jgresham

[jgresham]

we could just add on the checkpoint sync flag for each network config like here https://github.com/NiceNode/controller-api/blob/main/src/app/api/controller/NodeSpecs/lodestar/lodestar-v1.0.0.json#L67

I went with option 2 and tested it out in nice-node

it seems it's synchronising for Sepolia and Holesky for Consensus Clients.

Am I doing it right? @jgresham

It should be working. You could verify by comparing the latest block or slot shown in NiceNode or the logs of each client to a block explorer like https://sepolia.etherscan.io/ or https://sepolia.beaconcha.in/

[jgresham]

I suppose you should also see some confirmation of checkpoint sync working in the client logs when you start the node for the first time. After the node is synced it no longer uses checkpoint sync

[arionrefat]

Alright, I will check it out. Thank you

[arionrefat]

Yeah it seems to be it's working, the nodes are running. Where do I sent the PR then? https://github.com/NiceNode/controller-api here?

[jgresham]

Yeah it seems to be it's working, the nodes are running. Where do I sent the PR then? https://github.com/NiceNode/controller-api here?

yeah send the PR there please!

[arionrefat]

@jgresham Here is the PR : NiceNode/controller-api#18