Tutorial: JSON Basics¶
Machine data can be generated by a variety of devices, such as servers, cell phones, browsers, and so on. Messages sent from a device are called events. An event describes any single user action or occurrence that you want to track. Machines that collect large numbers of events may organize them into batches. A batch is a container that includes header information common to all of the events; e.g. source device and user information.
In a common data collection scenario, a scalable web endpoint collects POSTed data from different sources and writes them to a queuing system such as Amazon Kinesis, Apache Kafka, or RabbitMQ. An ingest service/utility then writes the data to a S3 bucket, from which you can load the data into Snowflake.
In this tutorial, you will learn how to partition JSON data batches in your S3 bucket, execute basic queries on loaded JSON data, and optionally flatten (removing the nesting from) repeated values. The actual data collection process is outside the scope of this tutorial.
|Interface:||SnowSQL (CLI Client)|
|Est. Time:||30-40 minutes|
|Est. Credits:||< 1|
Before You Begin: