AWS S3
Radar can stream both events and locations to AWS S3.
Use the AWS S3 integration to make Radar data available in a simple object storage format within your own systems, making your locations and events data more easily accessible by your engineers and data scientists.
By having all of your event and location data in S3, you can integrate your location data with other analysis tools like AWS Athena or Snowflake. Files are delivered via an AWS Kinesis Data Firehose as snappy-compressed parquet files. Each /dt=YYYY-MM-DD/ folder will have multiple files in it. This format should be consumable out-of-the-box with Apache Spark, AWS Athena, or other data engineering tools. This can enable you to build custom dashboard visualizations of data like foot traffic, trip duration, or dwell time in a location.
Configuration#
Contact your customer success manager to enable this feature.
The AWS S3 integration uses an AWS Kinesis Data Firehose to stream new Radar locations and events. Additionally, AWS KMS can be used to encrypt Kinesis data streams server-side.
Create a bucket#
Create an S3 bucket for Radar to deliver data to, for example, s3://com.example.bucket/. Data will be delivered to s3://com.example.bucket/radar/events/project=$PROJECT/dt=YYYY-MM-DD/X.parquet.snappy. Provide the name of the bucket to your customer success manager.
Create an AWS Cross-Account IAM policy#
Create an IAM Policy that allows Radar to deliver data to this bucket. To do this, implement step 2 in the Cross-Account Delivery to an Amazon S3 Destination IAM Policy documentation. Radar will implement steps 1 and 3. Contact your Customer Support Manager to obtain the values for the accountA-id and iam-role-name values to use in the IAM Policy.