对数据精度敏感的服务端不应该使用Infura

Infura的优点

在用Web3做以太坊相关的数据交互时候，需要节点来进行支撑，但是维护一个节点是非常占用资源和精力的；所以使用Infura这类开放平台的接口还是非常不错的；

Infura的缺点

接口限速

因为是开放的API，所以限速也不算缺点，可以通过充钱进行解决

接口数据精度不保证

假设用户在昨天（2019-12-11）拿到的数据A，今天（2019-12-12）拿到的数据可能是B

我在实际开发的过程中，遇到好多次这种情况；

其中一个例子：某个Token转账，通过Infura当时拿到的值是2000000000000000000，但是隔了N天以后，我拿到的数据是2000000000000000012，两个时间的值不一样；

所在HASH0x6ba281ff82f3e2ae5ce3b870e4cb712de99e1b629de2126afb0d47aa82aaa709;

官方浏览器 https://etherscan.io/tx/0x6ba281ff82f3e2ae5ce3b870e4cb712de99e1b629de2126afb0d47aa82aaa709 显示的数据是正确的；

也有几个第三方浏览器，比如 https://eth.tokenview.com/cn/tx/8869537/149 显示是2000000000000000000,猜测可能用的也是 Infura 接口；

现在的数据解析

现在已经是正确的数据了，下面只是一个解析的演示

通过下面RPC参数获取数据

{
    "jsonrpc":"2.0",
    "method":"eth_getLogs",
    "params":[
    {
    "address": "0x9eb7006b8b9010a9856f20611af462f872f6a40e",
    "fromBlock": "0x8756a1",
    "toBlock":"0x8756a1",
        "topics":["0x94fcee0b7b95ac21ec59ec2c5b2e99e75c909351baf99e93cd97e713d820627b"]
    }
    ],
    "id":1
 }

拿到的结果是

{
   "jsonrpc": "2.0",
   "id": 1,
   "result": [
       {
           "address": "0x9eb7006b8b9010a9856f20611af462f872f6a40e",
           "blockHash": "0x31ec3fc8987a2ae35b305f5aa837738e7253e82c7df10102119545a81861c206",
           "blockNumber": "0x8756a1",
           "data": "0x00000000000000000000000000000000000000000000000000000000000000600000000000000000000000001a762727adcfd0bf8895e5e10b9193269b89a7c5000000000000000000000000000000000000000000000000000000005dbfb58700000000000000000000000000000000000000000000000000000000000000202d1de569c149af4de8380d0792a1dc2a814377ca869ed3e12531a9cb4dbd0eaa",
           "logIndex": "0x8a",
           "removed": false,
           "topics": [
               "0x94fcee0b7b95ac21ec59ec2c5b2e99e75c909351baf99e93cd97e713d820627b",
               "0x0000000000000000000000000000000000000000000000001bc16d674ec8000c"
           ],
           "transactionHash": "0x6ba281ff82f3e2ae5ce3b870e4cb712de99e1b629de2126afb0d47aa82aaa709",
           "transactionIndex": "0x95"
       }
   ]
}

代码解析阶段

这里是拿到的数据解析

(async () => {
    const Web3 = require("web3")
    let config = {
        INFURA_URL: "https://rinkeby.infura.io/you_infura_key",
        INIT_BLOCK_NUM: 4655753,
        CONTRACT_ADDRESS: "0xA766d355Ef29502Ca68637F0BCb9Ff7284D1ace2"
    }

    const web3 = new Web3(config.INFURA_URL, { timeout: 1000 * 10 });
    const ABI = [
        {
            type: 'bytes',
            name: 'czrAccount'
        },
        {
            type: 'address',
            name: 'ethAddress'
        },
        {
            type: 'uint256',
            name: 'timestamp'
        },
        {
            type: 'uint256',
            name: 'value',
            indexed: true
        }
    ];

    let item = {
        "address": "0x9eb7006b8b9010a9856f20611af462f872f6a40e",
        "blockHash": "0x31ec3fc8987a2ae35b305f5aa837738e7253e82c7df10102119545a81861c206",
        "blockNumber": "0x8756a1",
        "data": "0x00000000000000000000000000000000000000000000000000000000000000600000000000000000000000001a762727adcfd0bf8895e5e10b9193269b89a7c5000000000000000000000000000000000000000000000000000000005dbfb58700000000000000000000000000000000000000000000000000000000000000202d1de569c149af4de8380d0792a1dc2a814377ca869ed3e12531a9cb4dbd0eaa",
        "logIndex": "0x8a",
        "removed": false,
        "topics": [
            "0x94fcee0b7b95ac21ec59ec2c5b2e99e75c909351baf99e93cd97e713d820627b",
            "0x0000000000000000000000000000000000000000000000001bc16d674ec8000c"
        ],
        "transactionHash": "0x6ba281ff82f3e2ae5ce3b870e4cb712de99e1b629de2126afb0d47aa82aaa709",
        "transactionIndex": "0x95"
    };
    let decodeLogInfo = await web3.eth.abi.decodeLog(
        ABI,
        item.data,
        [item.topics[1]]
    );
    let insertLog = {
        timestamp: decodeLogInfo.timestamp.toString(10),
        eth_hash: item.transactionHash,
        eth_address: decodeLogInfo.ethAddress,
        block_number: item.blockNumber,
        value: decodeLogInfo.value.toString(10),
        status: 1
    }
    console.log(insertLog)
})()

Infura的使用场景

这种的场景是用于前端的，比如网页，Chrome扩展（Matemask那种），轻钱包等等；这些会实时的获取数据，对精度不是那么的敏感；

Infura不适合的场景

不适合对精度要求非常敏感的服务端