ATLAS

ATLAS takes your Swagger, and smartly generates the code which could be used as input to various load testing tools. Currently, ATLAS fully supports Artillery Load Runner

ATLAS stands for Automation Tool for Load Analysis and Simulation

Creating a new Project

• Create a virtual environment. We only support Python 3.6+.
• Install the Atlas Project by pip install swagger-atlas
• Run atlas newproject <project_name>
• Switch to new directory
• Follow the README of directory to customize and run your load test

Features

What ATLAS Does

• Converts Swagger to Load Test file seamlessly
• Generates fake Data for testing.
• Understands resource dependencies and workflow - eg: if one API creates PET, and another API wants to use Pet ID, ATLAS is intelligent enough to:
  • Order the APIs so that CREATE API comes before any API which wants to use the IDs
  • Transfer the resource dependency
  • This works with arbitrarily with any complex relationships between APIs as long it is not cyclic
• Generate beautiful graphics and statistics

Do Read Use Cases for use cases and best swagger practices to get the most out of ATLAS Automation systems.

One of the many graphs which would be auto-generated:

---

Authenticating APIs

It may be possible that you need to provide Authentication Information with APIs. With ATLAS, you can provide Header Authentication by:

• Editing conf/profiles.yaml and adding authentication key value pairs
• In conf/artillery/hooks.js change addHeaders() function as needed.

Example

# profiles.yaml
<profile_name>:
    token: abcd

// hooks.js
async function addHeaders(profile) {
    profile.auth = {
        "headers": {'Authorization': 'Token ' + profile.token}
    };
    return profile;
}

This header will now be added to all API Requests. We only support Header authentication for now

Load Test Configuration

You should be able to edit load test configuration by following respective runner. For Artillery see CLI Options

Configuring ATLAS

ATLAS is highly automated system, however you may want to configure it to suit your needs. A quick overview of most frequent configuration options:

Manual Data Generation

ATLAS has an inbuilt AI which can generate fake data for most of requests. ATLAS can also extract data from one API and use it in another.

You can supply your own data for some of the resources. See Mapping Resources to Database section in Resources for details

---

Selective API Hits

You may want to load test a subset of APIs only. There are several quick ways to do that

• Exclude URLs: In settings, you can give a list of OP_KEYS for excluding URLs
• Tag Specific: (tags are the same as those given in swagger). See Profiles for tagging example
  • In settings, mark ONLY_TAG_API as True
  • In conf/profiles.yaml, in the tags section, mark the tags you want to test with
• You can generate your own scenarios. More details in Scenarios section.

---

Manipulating Requests

ATLAS allows you to manipulate request to API. You may want to do this to:

• Change the body being sent to an API
• Change URL being hit (eg: changing Query Params in it)
• Change Other Parameters (eg: Headers)

See Hooks for details on how to do this, and various examples for same

---

API Ordering

If you wanted to over-write ATLAS AI API ordering, you can do so. In conf/conf.py, change SWAGGER_OPERATION_DEPENDENCIES.

This setting is a list of 2-element tuple where it ensures that first element is always hit before second element. Each element represents OP_KEY of Swagger Operation. OP_KEY is simply "METHOD url" for any Swagger Operation OP_KEY for all swaggers can be obtained in conf/routes.py which is generated via atlas generate_routes

---

Generate Custom Scenarios

By default, ATLAS will hit all APIs in an order pre-determined by its AI. You may wish to specify a custom workflow or scenario.

ATLAS Provides such ability

• Writing custom scenario and linking it to profiles
• Profiles can have multiple scenarios associated with them
• If needed, can over-write default scenario

See Creating Custom Scenario section in Use cases for more details

Example Project

• After installing ATLAS, run atlas example <example_name> (To know possible examples you can run python atlas example --help)
• Switch to Project, and you will steps to run Server and Load Test

Contributing to ATLAS

Please see Contribution Guide