Metrics and Logging in API Manager

ColdFusion API Manager collects and analyzes information that APIs generate. This information is useful to measure the performance of APIs. The information provided by the API Manager can be used to make decisions, such as:

• Change in any API.
• Applications that are consuming an API for the maximum or minimum time.
• HTTP methods (GET, PUT, POST, and so on) that have maximum usage.
• Capacity planning of APIs.

Metrics

When an API makes a runtime request to the API Manager, the analytic server pulls the following information from the request, such as:

• Timestamp
• API Name
• Version
• Publisher
• Consumer
• Application
• SLA
• Error
• Request Method
• Status Code
• Request Content Size
• Response Content Size
• Request Time
• Sub resource pattern
• Request Content-Type
• Response Content-Type
• Cluster Node Name
• Request IP

To view the analytics, a publisher, consumer, or an administrator has access to a dashboard. Publishers and administrators have their own personal dashboards which they can edit and add a visualization, dashboard, and so on. But dashboards for consumers cannot be personalized.

This dashboard is a web interface that publishes metrics analytics and log information in the form of graphs and pie charts. 

Modify metric settings

As an administrator, you can configure the settings for the number of API requests for which to generate metrics. On the Administrator portal, select Server  > API Analytics Server Configuration to display the Metric Settings page.

Cluster Name- Enter the name of the cluster in which the APIs make the request. The analytics of these requests are then published into a portal. For more details, refer to Setting up Cluster Support.

Sniff-In Analytics Server configuration, the client can sniff the rest of the cluster, which adds nodes that the cluster can use. To enable the feature, select Sniff.

Flush Interval- Enter the time interval, in seconds, after which the ElasticSearch server receives a specified number of API requests.

Maximum Actions Per Bulk Request- Enter the number of API requests that are collectively sent to the ElasticSearch server for metrics calculation and visualization. If you enter 1000 in the field, no more than 1000 requests can be sent to the server.

Maximum Concurrent Bulk Request- Enter the number of concurrent blocks of API requests that are sent to the ElasticSearch server. For example, if you enter 2 in the field, no more than two concurrent blocks of API requests can be sent to the server.

Maximum Volume Per Bulk Request- Enter the size of the concurrent requests in this field. The size is calculated in MBs. For example, if you enter 5 in the field, the size of concurrent API requests cannot exceed 5 MB.

Log analysis in API Manager

A user (publisher or administrator or subscriber) can view the information that the API Manager logs.

Since the API Manager is a multi-tenant system, multiple publishers can use platform concurrently. So when a publisher retrieves the logged information, the API Manager retrieves those logs that are specific to a publisher. It is more secure since first publisher has no access to the second publisher's logs.

On the dashboard, the publisher can view and analyze logs using certain filters, such as:

• Logs in last 15 mins/30 mins/last day/last month/any custom time
• Error logs
• Logs from a specific consumer
• Logs for a particular service or a version
• A combination of one or more filters