AWS CloudWatch

Prerequisites

• Data source allows access to LogicLoop if necessary

• Deployment works for organization and data source

• LogicLoop user account is an admin to be able to create a data source

Setup

LogicLoop connects to CloudWatch using the boto3 client. According to their documentation before you can query with LogicLoop:

[Y]ou should set up authentication credentials. Credentials for your AWS account can be found in the IAM Console. You can create or use an existing user. Go to manage access keys and generate a new set of keys.

With your access_key_id and secret_access_key you can set up the CloudWatch and CloudWatch Log Insights data sources.

Querying CloudWatch

These instructions assume you are familiar with the CloudWatch ad-hoc query language. To make exploring your data easier the schema browser will show which Namespaces and Metrics you can query.

Your query text must written in YAML syntax (not SQL). It should be an array of MetricDataQuery objects under a key called MetricsDataQueries. When you execute the query LogicLoop converts this array to one or more get_metric_data() calls.

Here’s an example that sends two MetricDataQuery objects.

MetricDataQueries: 
  - Id: database_connections
    MetricStat:
      Metric:
        Namespace: AWS/RDS
        MetricName: DatabaseConnections
      Period: 60
      Stat: Maximum
  - Id: swap_usage
    MetricStat:
      Metric:
        Namespace: AWS/RDS
        MetricName: SwapUsage
      Period: 60
      Stat: Maximum
StartTime: "2020-01-01 00:00:00"

To query your CloudWatch data with SQL, first pull data using the YAML syntax. Then use the Query Results datasource to filter and sort the results

Helpful Links

To learn more about the CloudWatch ad-hoc query language you can study the following links.

• boto3 reference

• GetMetricData API Description

• MetricDataQuery Format

• MetricStat Format

Querying CloudWatch Logs Insights

These instructions assume you are familiar with the CloudWatch Logs Insights ad-hoc query language. To make exploring your data easier the schema browser will show which Log Groups and @Fields you can query. Your query text must be written in YAML Syntax (not SQL). Your query can include the following keys:

Key	Type
logGroupName	string
logGroupNames	array of strings
startTime	integer or timestring
endTime	integer or timestring
queryString	string
limit	integer

Your query must include either a logGroupName or logGroupNames. When you execute the query LogicLoop converts this object into a start_query() call. Here is an example query:

startTime: '2020-01-01 00:00'
logGroupNames: ['/ecs/rq-adhoc', '/ecs/rq-scheduled', '/ecs/rq-workers-scheduled']
queryString: 
  filter @message like 'Opened tunnel for <some org id>'
  | parse 'Opened tunnel for <some org id> (<data source id>) with port *' as port
  | fields @timestamp, @logStream
  | sort @timestamp desc

Helpful Links

• Query Syntax.