Accessing data from data lakes

Learn how to access data stored in data lakes.

info

If you want to try this example, you can deploy the following endpoint:

Read dta from S3 buckets across separate AWS accounts

Learn how to serve data live from two S3 buckets across two separate AWS accounts.

Overview

Usage:

/aws/s3/buckets

Code

main() =
    let
        // The type of each JSON file
        json_type = type record(creation_date: string, entries: list(record(hostname: string))),
        // Function returning the concatenated list of hostnames for a specific path & aws_config
        read_logs(path: string, aws_config: record(region: string, accessKey: string, secret: string)) =
            let
                // list all files of the bucket path
                bucket = S3.Build(
                    path,
                    region = aws_config.region,
                    accessKey = aws_config.accessKey,
                    secretKey = aws_config.secret
                ),
                files = Location.Ls(bucket),
                // open each file as JSON
                contents = List.Transform(files, (f) -> Json.Read(f, json_type)),
                // `Explode` the entries field
                entries = List.Explode(contents, (c) -> c.entries)
            in
                // project only the 'hostname' column to obtain the expected list of strings
                entries.hostname
    in
        let
            awsAccountA = {region: "eu-west-1", accessKey: "AKIAZ6SK5NCTDAAESLXU", secret: "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"},
            awsAccountB = {region: "eu-west-1", accessKey: "AKIAZ6SK5NCTLPK7QE4N", secret: "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"}
        in
            // Union the lists returned by `read_logs` for both buckets/accounts.
            List.Union(
                read_logs("s3://log-server-a/*.json", awsAccountA),
                read_logs("s3://log-server-b/*.json", awsAccountB)
            )

Introduction

Snapi supports reading data directly from S3.

To do so, you must first build a location and then read the file. For example, the following code reads a CSV file from S3:

let
    location = S3.Build("s3://my-bucket/data.csv", accessKey = "<AWS ACCESS KEY>", secretKey = "<AWS SECRET KEY>")
in
    Csv.InferAndRead(location)

You can use wildcards in the S3 location.

note

If you want to access data from other data lakes that are not yet supported, reach out to us.