norbybaru/laravel-aws-timestream

Library to interact with AWS Timestream service via API

v0.4.0 2024-04-02 23:37 UTC

This package is auto-updated.

Last update: 2024-04-04 08:45:05 UTC


README

PHPUnit Tests PHPStan

AWS Timestream is a fast, scalable, and serverless time series database service. This package is an opinionated implementation to query timestream and ingest data into timestream.

Upgrading from version 0.2.x

Please not that version 0.3.x is still backward compatible with version 0.2.x. No breaking changes introduced, however i will suggest you slowly use the new Payload builder approach for Timestream ingestion as from version 0.4.x we shall drop support for legacy builder.

See updated examples below start using new approach with TimestreamPayloadBuilder.

Install

composer require norbybaru/laravel-aws-timestream

Configuration

  • Publish config
php artisan vendor:publish --provider="NorbyBaru\AwsTimestream\TimestreamServiceProvider" --tag="timestream-config"
  • Open timestream.php config file and setup your database name and tables
  • Setup you AWS Timestream keys and permissions with the following environment variable
AWS_TIMESTREAM_KEY=
AWS_TIMESTREAM_SECRET=
AWS_TIMESTREAM_PROFILE=

Basic Usage

Query Timestream

Using TimestreamBuilder::query() will give autocomplete of all available functions

  1. Using TimestreamBuilder to build query to be passed onto TimestreamReaderDto which generate am object that can be consumed by TimestreamService query function
<?php

use NorbyBaru\AwsTimestream\TimestreamService;
use NorbyBaru\AwsTimestream\TimestreamBuilder;
use NorbyBaru\AwsTimestream\Dto\TimestreamReaderDto;

public function overview(TimestreamService $timestreamService)
{
    $queryBuilder = TimestreamBuilder::query()
        ->select('*')
        ->from("database-name", 'table-name')
        ->whereAgo('time', '24h', '>=')
        ->whereNotIn('measure_value::varchar', ['reviewer', 'open', 'closed'])
        ->orderBy('time', 'desc');
    
    TimestreamReaderDto::make($queryBuilder);

    // response from Aws timestream
    return $timestreamService->query($timestreamReader)
}
  1. Use TimestreamReaderDto to inject from query with default database name and on demand table name. NB. No need to add ->from() query on your query builder.
<?php

use NorbyBaru\AwsTimestream\TimestreamService;
use NorbyBaru\AwsTimestream\TimestreamBuilder;
use NorbyBaru\AwsTimestream\Dto\TimestreamReaderDto;

public function overview(TimestreamService $timestreamService)
{
   $queryBuilder = TimestreamBuilder::query()
       ->select('*')
       ->whereAgo('time', '24h', '>=')
       ->whereNotIn('measure_value::varchar', ['reviewer', 'open', 'closed'])
       ->orderBy('time', 'desc');
   
   TimestreamReaderDto::make($queryBuilder, 'table-name');

   // response from Aws timestream
   return $timestreamService->query($timestreamReader)
}

Timestream Ingestion

We need to build our ingestion payload that Timestream will accept for ingestion. To achieve that we can use the TimestreamPayloadBuilder class to build a payload that Aws Timestream will understand

Example

  1. Build single record ingestion payload
<?php

use NorbyBaru\AwsTimestream\TimestreamService;
use NorbyBaru\AwsTimestream\Dto\TimestreamWriterDto;
use NorbyBaru\AwsTimestream\Builder\TimestreamPayloadBuilder;

public function ingest(TimestreamService $timestreamService)
{
    $payload = TimestreamPayloadBuilder::make(measureName: 'cpu_usage')
        ->setMeasureValue(value: 80.5)
        ->setMeasureValueType(type: ValueTypeEnum::DOUBLE())
        ->setDimensions(name: "mac_address", value: '00:11:22:AA:BB:CC ')
        ->setDimensions(name: "ref", value: 'station a')
        ->setTime(Carbon::now());

    $timestreamWriter = TimestreamWriterDto::make($payload->toRecords())->forTable('table-name');

    return $timestreamService->write($timestreamWriter);
}
  1. Ingestion data in batch using Common Attributes to reduce ingestion cost with Timestream
<?php

use NorbyBaru\AwsTimestream\TimestreamService;
use NorbyBaru\AwsTimestream\Dto\TimestreamWriterDto;
use NorbyBaru\AwsTimestream\Builder\TimestreamPayloadBuilder;

public function ingest(TimestreamService $timestreamService)
{
    $payloads = [
        ...TimestreamPayloadBuilder::make(measureName: 'cpu_usage')
            ->setMeasureValue(value: 80.6)
            ->setDimensions(name: "ref", value: 'station a')
            ->toRecords(),
        ...TimestreamPayloadBuilder::make(measureName: 'memory_usage')
            ->setMeasureValue(value: 45.5)
            ->setDimensions(name: "ref", value: 'station b')
            ->toRecords(),
    ];

    $common = CommonPayloadBuilder::make()
        ->setCommonDimensions(name: 'processor', value: 'unix')
        ->setCommonDimensions(name: 'mac_address', value: '00:11:22:AA:BB:CC')
        ->setCommonDimensions(name: 'device_name', value: 'device_1')
        ->setCommonMeasureValueType(ValueTypeEnum::DOUBLE())
        ->setCommonTime(Carbon::now())
        ->toArray();

    $timestreamWriter = TimestreamWriterDto::make($payloads, $common, 'table-name');

    return $timestreamService->write($timestreamWriter);
}

Run Unit Test

composer test

Online Resources