Amazon DynamoDB

Official Documentation

The application stores recorded information in DynamoDB. Entries are partitioned by date and ordered by minute in day. The data is written by a simple Lambda which receives raw JSON from an IoT device topic.

The data is formatted as below:

{
  "caldate": "27-08-2018",
  "calminute": 4,
  "msg-id": "c7f07595-a4a2-4442-af90-2003282efe16",
  "pihum0": "89.57",
  "pipress0": "997.54",
  "pitemp0": "16.27"
}

The fields in the JSON are defined as:

Field	Definition
caldate	Date of data capture, format DD-MM-YYYY
calminute	Minute of capture day, range 0..1440
msg-id	Unique message identifier, UUID
pihum0	Humidity value, % range 0.0..100.00
pipress0	Pressure value, range 300..1100 hPa)
pitemp0	Temperatue value, range -40…85°C

The table to hold the IoT data is defined as a simple key/value mapping with the following structure:

Details	Value
Table Name	aws-iot-java
Primary partition key	caldate (String)
Primary sort key	calminute (Number)

This allows the data to be selected on a day to day basis, ordered by capture time.

Creation Script

Command: aws dynamodb create-table --table-name <schema file name>.json

Schema file:

{
   "AttributeDefinitions":[
      {
         "AttributeName":"caldate",
         "AttributeType":"S"
      },
      {
         "AttributeName":"calminute",
         "AttributeType":"N"
      }
   ],
   "KeySchema":[
      {
         "KeyType":"HASH",
         "AttributeName":"caldate"
      },
      {
         "KeyType":"RANGE",
         "AttributeName":"calminute"
      }
   ],
   "ProvisionedThroughput":{
      "WriteCapacityUnits":5,
      "ReadCapacityUnits":5
   },
   "TableName":"aws-iot-java"
}