javascript - AWS Lambda 函数多次处理相同的 dynamodb 流。我错过了什么？

Question

我编写了一个 node.js lambda 函数，该函数在将新记录插入特定表时基于 dynamodb 流触发。

该函数只接收新事件，过滤插入的记录，然后对于每条记录，使用几个字段从其他表中检索数据。使用此组合数据，可以编写消息并通过 SNS 将消息发送到特定目标 ARN。

该功能正确执行。检索所有相关数据，并发送推送通知。

但是，由于某种原因，该函数似乎为同一个流多次调用，并多次处理新插入的记录。结果是目标设备多次收到相同的推送通知。

我应该将回调放在不同的地方，还是我没有正确调用上下文？

这是功能:

'use strict';

var AWS = require("aws-sdk");
var dynamodb = new AWS.DynamoDB();
var sns = new AWS.SNS();

console.log('Loading function');

exports.handler = (event, context, callback) => {
  console.log('Received event:', JSON.stringify(event, null, 2));

  event.Records.forEach((record) => {
    console.log(record.eventID);
    console.log(record.eventName);
    console.log('DynamoDB Record: %j', record.dynamodb);

    if (record.eventName == 'INSERT') {
      var matchId = record.dynamodb.NewImage.eventId.S;
      var match_params = {
        Key: {
          "eventId": {
            S: matchId
          }
        },
        TableName: "xxxxxxxxxxx-mobilehub-xxxxxxx-Event"
      };

      //retrieve the match information from Event table
      dynamodb.getItem(match_params, function(err, data) {
        var match_description = "";
        if (err) {
          console.log(err, err.stack);
          context.fail('No match event record found in Event table');
        } else {
          match_description = data.Item.description.S;

          var uId = record.dynamodb.NewImage.participantUserId.S; //participantUserId 
          var user_params = {
            Key: {
              "userId": {
                S: uId
              }
            },
            TableName: "xxxxxxxxxxx-mobilehub-xxxxxxxxx-User"
          };

          //retrieve the user record from User table
          dynamodb.getItem(user_params, function(err, data) {
            if (err) {
              console.log(err, err.stack); // an error occurred  
              context.fail('Error occurred. See log.');
            } else {
              console.log(data); // successful response
              if (data.length === 0) {
                console.log("No User Record Found.");
                context.fail('No user found for participantUserId.');

              } else {

                var deviceARN = data.Item.device_arn.S;
                if (deviceARN <= 1) {
                  console.log("User has not registered their device for push notifications.");
                  context.fail('User has not registered for notifications');
                } else {

                  var json_message = JSON.stringify({
                    APNS_SANDBOX: JSON.stringify({
                      aps: {
                        alert: "You are playing in an upcoming match " + match_description,
                        badge: 1,
                        sound: 'default'
                      }
                    })
                  });

                  var snsparams = {
                    Message: json_message,
                    MessageStructure: 'json',
                    TargetArn: deviceARN
                  };

                  sns.publish(snsparams, function(err, data) {
                    if (err) {
                      console.log(err); // an error occurred
                      context.fail('SNS send failed. See log.');
                    } else {
                      console.log(data); // successful response
                      context.success('Push notification sent to user.');
                    }
                  });
                }
              }
            }
          });
        }
      });
    }
  });
  callback(null, `Successfully processed ${event.Records.length} records.`);
};

Answer 1

就我而言，我多次添加了相同的事件源。
引用与 AWS 支持工程师的对话:

Using my internal tools, I noticed that the Lambda function xxxxxx hasthe event source:arn:aws:events:my_region:my_acct_id:rule/my_event_targetconfigured twice as push event source. This means that this might be the causewhy you are seeing two invokes at every minute. Would you pleaseconfirm on your side if this event is configured twice for the $LATESTversion of your lambda and also confirm if it's intended?

我希望这可以节省一些:)