Add documentation, secure routes, etc

app/controllers/chaos_controller.rb

@@ -1,6 +1,8 @@
 # frozen_string_literal: true
 
 class ChaosController < ActionController::Base
+  before_action :validate_request
+
   def leakmem
     memory_mb = params[:memory_mb] ? params[:memory_mb].to_i : 100
     retainer = []
@@ -29,4 +31,15 @@ class ChaosController < ActionController::Base
     Process.kill("KILL", Process.pid)
   end
 
+  private
+
+  def validate_request
+    secret = ENV['GITLAB_CHAOS_SECRET']
+    return unless secret
+
+    unless request.headers["HTTP_X_CHAOS_SECRET"] == secret
+      render text: "To experience chaos, please set X-Chaos-Secret header", content_type: 'text/plain', status: 401
+    end
+  end
+
 end

config/routes.rb

@@ -83,10 +83,12 @@ Rails.application.routes.draw do
     draw :operations
     draw :instance_statistics
 
-    get '/chaos/leakmem' => 'chaos#leakmem'
-    get '/chaos/cpuspin' => 'chaos#cpuspin'
-    get '/chaos/sleep' => 'chaos#sleep'
-    get '/chaos/kill' => 'chaos#kill'
+    if ENV['GITLAB_ENABLE_CHAOS_ENDPOINTS']
+      get '/chaos/leakmem' => 'chaos#leakmem'
+      get '/chaos/cpuspin' => 'chaos#cpuspin'
+      get '/chaos/sleep' => 'chaos#sleep'
+      get '/chaos/kill' => 'chaos#kill'
+    end
   end
 
   draw :api

doc/development/chaos_endpoints.md

@@ -0,0 +1,83 @@
+# Generating Chaos in a test GitLab instance
+
+As [Werner Vogels](https://twitter.com/Werner), the CTO at Amazon Web Services, famously put it, **Everything fails, all the time**.
+
+As a developer, it's as important to consider the failure modes in which your software will operate as much as normal operation. Doing so can mean the difference between a minor hiccup leading to a scattering of 500 errors experienced by a tiny fraction of users and a full site outage affect all users for an extended period.
+
+To paraphrase [Tolstoy](https://en.wikipedia.org/wiki/Anna_Karenina_principle), _all happy servers are alike, but all failing servers are failing in their own way_. Luckily, there are ways we can attempt to simulate these failure modes, and the chaos endpoints are tools for assisting in this process.
+
+Currently, there are four endpoints for simulating the following conditions: slow requests, cpu-bound requests, memory leaks and unexpected process crashes.
+
+## Enabling Chaos Endpoints
+
+For obvious reasons, these endpoints are not enabled by default. They can be enabled by setting the `GITLAB_ENABLE_CHAOS_ENDPOINTS` environment variable.
+
+For example, if you're using the [GDK](https://gitlab.com/gitlab-org/gitlab-development-kit) this can be done with the following command:
+
+```shell
+GITLAB_ENABLE_CHAOS_ENDPOINTS=1 gdk run
+```
+
+### Securing the Chaos Endpoints
+
+**It is highly recommended that you secure access to the Chaos endpoints using a secret token**. This is recommended when enabling these endpoints locally, and essential when running in a staging or other shared environment. _It goes without saying that you should not enable them in production unless you absolutely know what you're doing._
+
+A secret can be set through the `GITLAB_CHAOS_SECRET` environment variable. For example, when using the [GDK](https://gitlab.com/gitlab-org/gitlab-development-kit) this can be done with the following command line:
+
+```shell
+GITLAB_ENABLE_CHAOS_ENDPOINTS=1 GITLAB_CHAOS_SECRET=secret gdk run
+```
+
+Replace `secret` with your own secret token.
+
+## Invoking Chaos
+
+Once you have enabled the chaos endpoints and restarted the application you can start testing using the endpoints.
+
+### Memory Leaks
+
+To simulate a memory leak in your application, use the `/-/chaos/leakmem` endpoint.
+
+For example, if your GitLab instance is listening at `localhost:3000`, you could `curl` the endpoint as follows:
+
+```shell
+curl http://localhost:3000/-/chaos/leakmem?memory_mb=1024 -H 'X-Chaos-Secret: secret'
+```
+
+The `memory_mb` parameter tells the application how much memory it should leak.
+
+Note: the memory is not retained after the request, so once its completed, the Ruby garbage collector will attempt to recover the memory.
+
+### CPU Spin
+
+This endpoint attempts to fully utilise a single core, at 100%, for the given period.
+
+```shell
+curl http://localhost:3000/-/chaos/cpuspin?duration_s=60 -H 'X-Chaos-Secret: secret'
+```
+
+The `duration_s` parameter will configure how long the core is utilised.
+
+Depending on your rack server setup, your request may timeout after a predermined period (normally 60 seconds). If you're using Unicorn, this is done by killing the worker process.
+
+### Sleep
+
+This endpoint is similar to the CPU Spin endpoint but simulates off-processor activity, such backend services of IO. It will sleep for a given duration.
+
+```shell
+curl http://localhost:3000/-/chaos/sleep?duration_s=60 -H 'X-Chaos-Secret: secret'
+```
+
+The `duration_s` parameter will configure how long the request will sleep for.
+
+As with the CPU Spin endpoint, this may lead to your request timing out if duration exceeds the configured limit.
+
+### Kill
+
+This endpoint will simulate the unexpected death of a worker process using a `kill` signal.
+
+```shell
+curl http://localhost:3000/-/chaos/kill -H 'X-Chaos-Secret: secret'
+```
+
+Note: since this endpoint uses the `KILL` signal, the worker is not given a chance to cleanup or shutdown.

doc/development/performance.md

@@ -41,6 +41,7 @@ GitLab provides built-in tools to aid the process of improving performance:
 * [GitLab Performance Monitoring](../administration/monitoring/performance/index.md)
 * [Request Profiling](../administration/monitoring/performance/request_profiling.md)
 * [QueryRecoder](query_recorder.md) for preventing `N+1` regressions
+* [Chaos Endpoints](chaos_endpoints.md) less for performance, more for availability: tools for testing failure scenarios
 
 GitLab employees can use GitLab.com's performance monitoring systems located at
 <https://dashboards.gitlab.net>, this requires you to log in using your