扫描DynamoDB表未返回数据

Question

I am trying to scan failed token counts from dynamodb database table without any indexes. It is returning 0 from the database. I doubt it is not scanning complete database. Below is my method and the dynamoDBClient working condition one and it has connection details. I am posting here only the scan query part

public int getFailedAuthStatusCount() {

    Map<String,String> expressionAttributesNames = new HashMap<>();
      expressionAttributesNames.put("#status","auth_status");

    Map<String, AttributeValue> expressionAttributeValues = new HashMap<String, AttributeValue>();
    expressionAttributeValues.put(":val", new AttributeValue().withS("FAIL"));

    ScanRequest scanRequest = new ScanRequest()
              .withTableName("Token")
              .withFilterExpression("#status = :val")
              .withExpressionAttributeNames(expressionAttributesNames)
              .withExpressionAttributeValues(expressionAttributeValues);
    ScanResult scanResult = dynamoDBClient.scan(scanRequest); //client is working fine.
    return scanResult.getCount();
  }

Here is the response.

{Items: [],Count: 0,ScannedCount: 1456,LastEvaluatedKey: {GUID={S: 0c4b281e6f9290c0fb3bf13f28c88fd,}, VENDOR={S: DELL,}},}

what is wrong with my request?

Answer 1

Nothing is wrong with your request. When scanning, you have to continue resubmitting the request with the prior LastEvaluatedKey as the next request's ExclusiveStartKey until you no longer receive a LastEvaluatedKey in the response.

Your first request is only evaluating the first ScannedCount: 1456 items in the table (≈ 1MB of data).

See https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/Scan.html#Scan.Pagination

Table scans are slow and expensive in all database environments. That's a key reason why indexes are important. DynamoDB makes this more apparent because of the low-level API that's available.