eth_getLogs suggest event ranges which can never be fetched

[joshstevens19]

🐛 Bug Report

📝 Description

Hey, I am running the zkSync local node and hit an issue; the node supplies a message which suggests block ranges when requesting large amounts of logs, which frameworks use for indexing; all other providers like Alchemy or Tenderly do similar as well to allow you to skip over events faster instead of having to scan block by block:

Doing a request like:

{
  "jsonrpc":"2.0",
  "method":"eth_getLogs",
  "params":[{
      "fromBlock": "0x0",
       "topics":["0x3ae390be004f906708f14ce7e66ca2ce9e1ad7ba15cd6401d7479f74cc70a58e"]
  }],
  "id":74
}

would yield back:

{
    "jsonrpc": "2.0",
    "error": {
        "code": -32602,
        "message": "Query returned more than 10000 results. Try with this block range [0x0, 0x30f]."
    },
    "id": 74
}

which makes sense the indexing framework would extract that range and do the next request:

{
	"jsonrpc":"2.0",
	"method":"eth_getLogs",
	"params":[{
             "fromBlock": "0x0",
             "toBlock": "0x30f",
	     "topics":["0x3ae390be004f906708f14ce7e66ca2ce9e1ad7ba15cd6401d7479f74cc70a58e"]
	}],
	"id":74
}

but this then errors with:

{
    "jsonrpc": "2.0",
    "error": {
        "code": -32008,
        "message": "Response is too big",
        "data": "Exceeded max limit of 10485760"
    },
    "id": 74
}

This means indexing systems break as they expect the suggested range to always work, and indexing frameworks should not have to change the core code to handle this edge case.

🔄 Reproduction Steps

Find an event which has emitted over 10,000 events and produced a bigger payload than 10,485,760

🤔 Expected Behavior

The suggested blocks should consider the event payload response size to give a range which will always return from the node

😯 Current Behavior

The suggested blocks given back in eth_getLogs do not work meaning indexing systems would not be able to index that event