Mainnet changes

[svlachakis]

The point of this PR is to start Nitro from block 22million+ which is Arbitrum One first block, but without having users to load the 32GB pebble genesis state of nitro.

Refactored our custom env variables to config properties.

[hudem1]

For my understanding, so, what this essentially does is fetch the start block data from NMC and set it to some consensus db. And this is enough for nitro to pick up from that block and continue as usual, meaning fetch the next blocks from L1 to send to NMC for execution ?

If so, then NMC has to use its internal comparison mode to make sure the block hashes its computing are consistent with l1, right ?

[svlachakis]

For my understanding, so, what this essentially does is to fetch the start block data from NMC and set it to some consensus db. And this is enough for nitro to pick up from that block and continue as usual, meaning fetch the next blocks from L1 to send to NMC for execution ?

If so, then NMC has to use its internal comparison mode to make sure the block hashes its computing are consistent with l1, right ?

Yes, your understanding is correct. After bootstrap:

• Nitro reads batches from L1
• Sends them to Nethermind for execution
• Nethermind executes and it's the only execution engine - works in external mode not in compare
• Returns results to Nitro

[hudem1]

For my understanding, so, what this essentially does is to fetch the start block data from NMC and set it to some consensus db. And this is enough for nitro to pick up from that block and continue as usual, meaning fetch the next blocks from L1 to send to NMC for execution ?
If so, then NMC has to use its internal comparison mode to make sure the block hashes its computing are consistent with l1, right ?

Yes, your understanding is correct. After bootstrap:

• Nitro reads batches from L1
• Sends them to Nethermind for execution
• Nethermind executes and it's the only execution engine - works in external mode not in compare
• Returns results to Nitro
• Nitro validates and stores

Oh okay nitro validates ? validates what exactly ? i mean, the block hashes validation was done by the compare client implem

[svlachakis]

For my understanding, so, what this essentially does is to fetch the start block data from NMC and set it to some consensus db. And this is enough for nitro to pick up from that block and continue as usual, meaning fetch the next blocks from L1 to send to NMC for execution ?
If so, then NMC has to use its internal comparison mode to make sure the block hashes its computing are consistent with l1, right ?

Yes, your understanding is correct. After bootstrap:

• Nitro reads batches from L1
• Sends them to Nethermind for execution
• Nethermind executes and it's the only execution engine - works in external mode not in compare
• Returns results to Nitro
• Nitro validates and stores

Oh okay nitro validates ? validates what exactly ? i mean, the block hashes validation was done by the compare client implem

Nothing - I got confused. In external mode nitro doesn't validate it works only as consensus

[damian-orzechowski]

I think config at execution level should be more general - so ExternalUrl instead of NethermindUrl

[svlachakis]

@damian-orzechowski as discussed on slack merging as is and we can change them later if needed