Integrations for Sysdig Monitor

Integrations for Sysdig Monitor lets you monitor infrastructure, applications, and cloud accounts. Sysdig supports metrics from Prometheus, JMX, StatsD, Kubernetes, and a number of applications to provide a 360-degree view of your environment. Many metrics are collected as soon as the agent is installed.

The Integrations module of Sysdig Monitor features four Data Sources:

  • Monitoring Integrations: Manage the collection of metrics across multiple environments via Monitoring Integrations.
  • Cloud Accounts: Collect metrics from cloud providers such as AWS, GCP, and Azure.
  • Sysdig Platform Audit: Review audit logs of activity on Sysdig Platform. See Sysdig Platform Audit.
  • Sysdig Agents: Check the status of Sysdig agents. See Sysdig Agents.

This page covers the basic concepts behind Integrations.

What is an Integration?

An integration refers to the process of connecting Sysdig with other applications and platforms. An integration enables the Sysdig agent to scrape an application, such as Nginx, Kubelet, or HAProxy, for metrics. Integrations can also refer to cloud accounts, the Activity Audit, or the Sysdig Agents pages.

What is a Metric?

A metric is a logical collection of time series sharing the same name, such as workqueue_depth, kube_job_complete, or http_requests. Many applications make dozens of metrics available for collection. Sysdig Integrations prioritize collection of the most valuable metrics for use in modules such as Dashboards, and Alerts.

What is a Time Series?

A metric is identified by its unique name, such as http_requests. A time series is defined by both the metric name and a distinct set of label key-value pairs, for example http_requests{method="GET", endpoint="/api/v2/users"}. Every time series is unique, since no two can have an identical combination of labels.

MetricTimes SeriesValueDescription
http_requestshttp_requests{method="GET", endpoint="/api/v2/users"}100number of GET HTTP requests to /api/v2/users
http_requestshttp_requests{method="POST", endpoint="/api/v2/users"}250number of POST HTTP requests to /api/v2/users
http_requestshttp_requests{status="500", endpoint="/api/v2/users"}10Number of HTTP requests returning 500 to /api/v2/users

Default and Custom Metrics

Sysdig Agent collects thousand of metrics at no additional cost, such as sysdig_container_net_server_in_bytes or kube_pod_info; these are called default metrics.

On the other hand, there are custom metrics, which do count towards your Time Series entitlement limit. They can be identified with the _sysdig_custom_metric="true" label. Any metric without this label is a default metrics that doesn’t count towards your entitlement limit.

How Are Metrics Collected?

The Sysdig Agent will begin collecting and reporting metrics as soon as it is installed. See Install Sysdig Agent

For applications in environments where the agent is not installed, you can send metrics to Sysdig with Prometheus Remote Write. See Configure Prometheus Remote Write.

Cloud accounts start sending metrics as soon as they are connected to Sysdig. See Cloud Accounts.

Customize Metrics Collection

Once an integration is set up, the Sysdig agent will begin to collect metrics. To configure metrics collection:

  • Enable an integration to collect metrics from that integration. See Enable and Disable Integrations.
  • For metrics collected by the Sysdig Agent, you can exclude clusters from collection through the UI.
  • For metrics collected via Prometheus Remote Write, edit the prometheus.yaml file to exclude the collection of certain metrics.
  • Edit the annotation of a pod to prevent the collection of metrics by the Sysdig Agent. See Exclude Workloads from Collection.
  • For Cloud Account metrics, modify metrics collection from the Cloud Accounts module. See Understand Cloud Accounts UI.
  • See Reduce Metrics Consumption.

These methods let you add new metrics or drop metrics from being collected.