Manual Triggers

Motivation

The main use-case for manual triggers is making troubleshooting easier.

This ties into Robusta's goal of reducing MTTR (mean time to recovery).

Additional use cases are discussed below.

Troubleshooting examples

A few examples:

  1. Debug Python pods in VSCode

  2. Find memory leaks in applications

  3. Function-level CPU profiling

Other examples

Manual actions aren't just for troubleshooting. You can automate any repetitive task on Kubernetes:

  1. Run chaos engineering scenarios

  2. Stress test pods over HTTP

How it works

Internally, troubleshooting actions are implemented the same way as other Robusta actions, like insights and automated fixes. A manual action is simply an action that can be triggered manually using the CLI.

Many actions supports both manual and automated triggers.

How to use manual triggers

Use the Robusta CLI to manually trigger a supported action:

robusta playbooks trigger <action_name> name=<name> namespace=<namespace> kind=<kind> <key>=<value>

The parameters above are:

name

The name of a Kubernetes resource

namespace

The resource's namespace

kind

pod, deployment, or any other resource the action supports. This can be left out for playbooks that support one input type.

<key>=<value>

Any additional parameters the action needs