Figure 1 : IoT Standard Pattern with Datastore and CloudSQL
properly protected , can be accessed at high speed and provide the ability to query in ad-hoc ways by various parties who are interested in research and modeling with past data . The platform must provide capabilities to :
• Separate transactional data from read-intensive data and place each data type in an optimized platform with the appropriate access patterns
• Denormalize data in a way that ensures future changes to the data model are simplified
• Choose data platforms that use a common interface that software engineers are comfortable with , this is most often SQL for relational data sets
Data Storage
IoT platforms collect data with different transactional characteristics . The Google Cloud provides a variety of data storage engines to accommodate data of different performance requirements and enable both traditional data models and denormalized models . Common IoT platforms will leverage a combination of tools based on data type :
• Cloud SQL – CloudSQL provides a relationship environment with strict enforcement of relationships and data integrity . CloudSQL is recommended for all contextual data including user information , device information and transactional records that can require updates .
• DataStore – Google Datastore is a NoSQL platform , providing a SQL interface , but targeted for read-intensive data stored in wide tables . DataStore provides efficient storage for time series data and other events generated by devices that need to be read , but will not be updated once received .
• Cloud Storage – Google Cloud Storage is an efficient and scalable location to store objects , such as raw data from sensors and devices , or as an archive environment for data that ’ s not in a transactional environment but is still needed .
IoT platforms are about the seamless acquisition , transformation and analysis of data , empowering users to make more informed decisions . Today , users demand high quality analytical results that will provide them with valuable life information provided by their IoT platform . Google CloudSQL and Datastore ensure that user data is stored at high speeds , available for analysis and stored in a flexible way for adding capabilities and features as time goes on .
30 | THE DOPPLER | SPRING 2016