Orchestrate
The Reliably orchestrator is at the heart of the platform and your daily experience with Reliably.
Centralize experiments
Reliably experiments are akin to Chaos Engineering experiments and the core mechanism via which all the data collected by Reliably can be turned into relevant information for you.
Plan & Schedule experiments
Plan and schedule experiments so they run now or in a repeated fashion so you can follow their impact trend.
Deploy anywhere
Reliably offers a variety of ways for you to execute experiments, such as in Reliably’s Cloud where we manage the infrastructure for you. We also support running from your CI or pretty much anywhere you require by using the Reliably CLI.
Pause, Resume, Terminate executions
Reliably understands that experiment on a system is not always an automated activity.
Sometimes you need to let Reliably know that it should delay continuing the execution by pausing it. This gives the possibility to explore the system even more thoroughly while under the condition given by the experiment. Of course, no pause function should exist without its counterpart to resume the execution.
We also appreciate that you may need to interrupt the execution as soon as possible and let’s you do so manually or via automated safeguards.
Overall, Reliably gives you the control over executions at all times.
Lock-in free
Reliably core experiments are Chaos Toolkit experiments.
The Chaos Toolkit is the popular open-source Chaos Engineering tool for everyone. Reliably provides benefits on top of Chaos Toolkit experiments but does not bring any lock-in features on top of it.
version: 1.0.0
title: Latency remains under 200ms
description: Verify that our endpoint responds under a reasonable amount of time
tags:
- latency
contributions:
errors: high
latency: high
security: none
availability: low
runtime:
hypothesis:
strategy: after-method-only
configuration:
reliably_url:
key: RELIABLY_PARAM_URL
type: env
default: 'https://reliably.com'
reliably_latency:
key: RELIABLY_PARAM_LATENCY
type: env
default: 0.2
steady-state-hypothesis:
title: capture-response-time-and-verify-it
probes:
- name: measure-endpoint-response-time
type: probe
provider:
func: measure_response_time
type: python
module: chaosreliably.activities.http.probes
arguments:
url: '${reliably_url}'
tolerance:
name: validate-response-time
type: probe
provider:
func: response_time_must_be_under
type: python
module: chaosreliably.activities.http.tolerances
arguments:
latency: '${reliably_latency}'
method: []