Many types of logs are meant to be used for telemetry to track trends, such as KPIs, over long periods of time. Generating metrics from your logs is a cost-effective way to summarize log data from high-volume logs, such as CDN logs, VPC flow logs, firewall logs, and networks logs. Use the generate metrics processor to generate either a count metric of logs that match a query or a distribution metric of a numeric value contained in the logs, such as a request duration.

Note: The metrics generated are custom metrics and billed accordingly. See Custom Metrics Billing for more information.

To set up the processor:

Click Manage Metrics to create new metrics or edit existing metrics. This opens a side panel.

• If you have not created any metrics yet, enter the metric parameters as described in the Add a metric section to create a metric.
• If you have already created metrics, click on the metric’s row in the overview table to edit or delete it. Use the search bar to find a specific metric by its name, and then select the metric to edit or delete it. Click Add Metric to add another metric.

Add a metric

1. Enter a filter query. Only logs that match the specified filter query are processed. All logs, regardless of whether they match the filter query, are sent to the next step in the pipeline. Note: Since a single processor can generate multiple metrics, you can define a different filter query for each metric.
2. Enter a name for the metric.
3. In the Define parameters section, select the metric type (count, gauge, or distribution). See the Count metric example and Distribution metric example. Also see Metrics Types for more information.
   • For gauge and distribution metric types, select a log field which has a numeric (or parseable numeric string) value that is used for the value of the generated metric.
   • For the distribution metric type, the log field’s value can be an array of (parseable) numerics, which is used for the generated metric’s sample set.
   • The Group by field determines how the metric values are grouped together. For example, if you have hundreds of hosts spread across four regions, grouping by region allows you to graph one line for every region. The fields listed in the Group by setting are set as tags on the configured metric.
4. Click Add Metric.

Metrics Types

You can generate these types of metrics for your logs. See the Metrics Types and Distributions documentation for more details.

Count metric example

For this status:error log example:

{"status": "error", "env": "prod", "host": "ip-172-25-222-111.ec2.internal"}

To create a count metric that counts the number of logs that contain "status":"error" and groups them by env and host, enter the following information:

Distribution metric example

For this example of an API response log:

{
    "timestamp": "2018-10-15T17:01:33Z",
    "method": "GET",
    "status": 200,
    "request_body": "{"information"}",
    "response_time_seconds: 10
}

To create a distribution metric that measures the average time it takes for an API call to be made, enter the following information:

Filter query syntax

Each processor has a corresponding filter query in their fields. Processors only process logs that match their filter query. And for all processors except the filter processor, logs that do not match the query are sent to the next step of the pipeline. For the filter processor, logs that do not match the query are dropped.

For any attribute, tag, or key:value pair that is not a reserved attribute, your query must start with @. Conversely, to filter reserved attributes, you do not need to append @ in front of your filter query.

For example, to filter out and drop status:info logs, your filter can be set as NOT (status:info). To filter out and drop system-status:info, your filter must be set as NOT (@system-status:info).

Filter query examples:

• NOT (status:debug): This filters for only logs that do not have the status DEBUG.
• status:ok service:flask-web-app: This filters for all logs with the status OK from your flask-web-app service.
  • This query can also be written as: status:ok AND service:flask-web-app.
• host:COMP-A9JNGYK OR host:COMP-J58KAS: This filter query only matches logs from the labeled hosts.
• @user.status:inactive: This filters for logs with the status inactive nested under the user attribute.

Learn more about writing filter queries in Datadog’s Log Search Syntax.