neoload 0.4.2

A command-line native utility for launching and observing NeoLoad performance tests

NeoLoad CLI

This command-line interface helps you launch and observe performance tests on the Neotys Platform. Since NeoLoad is very flexible to many deployment models (SaaS, self-hosted, cloud or local containers, etc.), configuration and test execution parameters depend on your licensing and infrastructure provisioning options. Please read the following instructions carefully.

TL;DR ... What

The goal of this guide is to demonstrate how you can:

1. create API load tests using code (YAML)
2. run them from any environment
3. visualize test results in web dashboards

TL;DR ... How

pip3 install neoload
git clone https://github.com/Neotys-Labs/neoload-cli.git && cd neoload-cli
neoload --profile saas --token [NLW_TOKEN] --zone [NLW_ZONE_ID] \
        --attach docker#2,neotys/neoload-loadgenerator:latest \
        -f tests/example_2_0_runtime/default.yaml \
        --scenario sanityScenario

NOTE: For Windows command line, replace the '\' multi-line separators above with '^'

Contents

• Prerequisites
• Installation
• Configuration
• Running Load Tests
• Additional Options
  • Exporting SLA Results to JUnit
  • Test Summary
  • Test Modifications
• Continuous Testing Examples
• Packaging the CLI with Build Agents
• IDE Integrations
• Contributing

Prerequisites

The examples below assume that you have Python3 and Git command line tools installed.

For Windows 10 users, see:

• Python In short:
  • Just install via Python.org Downloads, then
  • Open a command prompt and install pip:

    python -m pip install -U pip
    

  • If the above doesn't work (Microsoft App Store sometimes confuses the situation), more details on installing Python in Windows can be found here.
• Git Command Line Tools See 5 Ways to Install Git on Windows
• Docker for Windows To install Docker with Chocolatey
  • Install Chocolatey package manager for Windows
  • Open a command prompt and install Docker Desktop for Windows 10

    choco install docker-cli docker-desktop
    

  • You probably will need to log out (or restart) Windows for Docker to work properly

For Mac OS X:

• Installing Python3 on Mac OS X
• From: Installing Git, the Easy Way

For packaging in Docker (custom build agents, not Docker for Windows), see appendix.

Installation

To install, simply run the following command. As of Jan 2020, Python 2 will be permanently deprecated, therefore this utility is written for Python 3.

pip3 install neoload

Interactive CLI Help

For usage and CLI argument examples, simply run:

neoload --help

Configuration

Before running a load test, you will need to initialize a profile to contain connection information to NeoLoad Web. Subsequent commands will use this profile unless switched to another profile, as there can be multiple. This is so that you only need to specify these details once only, then you can focus on validating and running load tests.

Your profile details depend on your license model and how you want to obtain load testing resources (controller + load generators). See below:

Create a SaaS-Based License Profile

If you have an enterprise license activated in NeoLoad Web SaaS, you do not need to specify license details in your profile. When a controller begins to run a test, it will reach out to NeoLoad Web SaaS to lease an appropriate license.

neoload --profile saas1 --token [NLW_TOKEN] --zone [NLW_ZONE_ID]

Create a Self-Hosted Enterprise License Profile

If you do not have a SaaS-based license, you will need to specify additional licensing server url and credentials that refer to your own NeoLoad Team License Server.

neoload --profile nts --token [NLW_TOKEN] --zone [NLW_ZONE_ID] --ntsurl [TEAM_SERVER_URL] --ntslogin [TEAM_SERVER_USERNAME_COLON_ENC_PASS]

Viewing saved profiles

You can always list profiles that have saved using the following command:

neoload --profiles

You can also view the raw/complete stored JSON representation of saved profile details using the following command:

neoload --profile [your_profile_name] --summary

Running Your First Load Test

Before proceeding, for now skip to the Obtain Basic Examples and run your first test to make sure everything is working as expected.

STOP HERE AND DO THE ABOVE BEFORE READING FURTHER

Future Plans to Execute Local Tests on a Free Trial License

If you do not already have an enterprise license, such as if you only have NeoLoad installed on your local workstation (from a Free Trial download or professional license), NeoLoad CLI will eventually support using your installation as an attached controller and load generator.

Additionally, if you do not have access to a dynamic infrastructure provider (i.e. OpenShift, etc.) and if you do not have Docker installed where the NeoLoad CLI is executed, your local workstation can serve as load testing resources similar to dynamic and containerized resource use cases.

This is planned to be delivered in Dec 2019.

Running Load Tests

Once a profile is established, NeoLoad CLI makes it very easy to execute load tests. All you need is to provide an existing test suite or set of as-code file(s), then specify a scenario.

NeoLoad CLI defaults to using the NeoLoad Web APIs for Runtime operations, which means that your project assets will be zipped up together and uploaded to the NeoLoad Web deployment you specified in your profile (SaaS or self-hosted).

neoload --scenario sanityScenario -f [path_to_your_nlp_or_yaml_file]

Once a test is initialized, if you are running in interactive console mode, the NeoLoad CLI will automatically open the system default browser to your live test results.

Customizing Test Name and Description

Though the default test name includes the scenario, you may want to customize the test name with specific details (e.g. CI job and build number, etc.). To do so add --testname, and optionally --testdesc to provide a readable test description:

neoload --scenario sanityScenario -f [path_to_your_nlp_or_yaml_file] \
        --testname "MyCustomTestName_${JOB_ID}" \
        --testdesc "A custom test description containing hashtags like #latest or #issueNum"

Obtain Basic Examples

Some basic examples are in our Git repository for this utility, under the directory ./tests/. To get them, simply clone the repo:

git clone https://github.com/Neotys-Labs/neoload-cli.git && neoload-cli

Then you can run a basic test simply by typing in:

neoload --attach docker#2,neotys/neoload-loadgenerator:latest
neoload -f tests/example_2_0_runtime/default.yaml --scenario sanityScenario

Additionally, you can specify multiple files, such as additional SLA, variables, or servers overriding files. This works for both [.nlp] and [.yaml] files.

neoload -f tests/example_2_0_runtime/default.yaml -f tests/example_2_0_runtime/slas/uat.yaml --scenario sanityScenario

Non-blocking Execution Workflow

You may want to manage individual steps of attaching resources, running, waiting, and detatching resources in a parallel pipeline. This is particularly useful in combination with other parallel steps that dynamically analyze SLA data and fast-fail the test (using API commands) using a known custom real-time analysis process.

The general process can be seen in the NeoLoad CLI E2E PyTest suite, but also abstracted below:

• Initialize a profile

  neoload --profile example --zone [static_or_dynamic_zone_id] --token [your_neoload_web_token]
  

• (Optional for static zone) Attach dynamic resources

  neoload --attach docker#2,neotys/neoload-loadgenerator:latest
  

  (Note: this assumes that the host you run the above command is a Docker host or is connected to one, DinD)

• Kick off a test via NeoLoad Web Runtime

  neoload --nowait --scenario [your_scenario_name] -f [your_project_file(s)] --outfile neoload.stdout
  

• Grab the ID of the test just executed

  neoload --infile neoload.stdout --query testid
  

• (Start some other parallel operations, such as monitoring for early advanced SLA failures)

• Wait (blocking) for test to complete

  neoload --spinwait --summary --junitsla junit_neoload_sla.xml --testid [test_id_obtained_from_above_query]
  

• (Always) Detatch Optional Docker resources

  neoload --detatch
  

Additional Options

There are many other arguments for test summarization, modification, and exporting results.

Using more than one Controller and Load Generator from a zone

Presuming that you already have a zone with available resources in it, you can specify to use more than one load generator. The zone should be stored in your current profile.

  neoload --attach zone#3 -f tests/example_2_0_runtime/default.yaml --scenario sanityScenario

...where the number three is the number of available load generators you want this test to utilize.

Exporting SLA Results to JUnit

  neoload --testid [guid] --junitsla=junit_sla_results.xml

Verbose and Debug Mode

Verbose mode changes logging level to INFO. When executing from an interactive console, it also opens a browser window to the test logs immediately after they're available (useful for monitoring the infrastructure initialization process).

neoload -f tests/example_2_0_runtime/default.yaml --scenario sanityScenario --verbose

Debug mode changes logging level to DEBUG. This is an extreme amount of internal information which infers verbose/INFO mode and provides manual pausing on critical events (such as after Docker attach but before test execution).

neoload -f tests/example_2_0_runtime/default.yaml --scenario sanityScenario --debug

Quiet Mode

Adding the --quiet flag generally produces only the relevant output (i.e. JSON, test id, etc.) and no other informational messages. This is useful when piping the structured data output into another process, such as obtaining the most recent test id from stdout file (--quiet is inferred by some other combinations of flags such as --infile AND --query testid).

Test Summary

Specifying both a --testid value and the --summary flag produces a JSON result set that shows both the high level test status metadata and overall statistics.

neoload --testid c9f9c994-da31-4b63-b622-42a80e313d15 --summary --quiet

Produces:

{
	'summary':{
		'author':'Paul Bruce',
		'description':'',
		'duration':10921,
		'end_date':1575675769036,
		'id':'c9f9c994-da31-4b63-b622-42a80e313d15',
		'lg_count':1,
		'name':'NeoLoad-CLI-example-2_0_sanityScenario',
		'project':'NeoLoad-CLI-example-2_0',
		'quality_status':'FAILED',
		'scenario':'sanityScenario',
		'start_date':1575675758115,
		'status':'TERMINATED',
		'termination_reason':'POLICY'
	},
	'statistics':{
		'last_request_count_per_second':None,
		'last_transaction_duration_average':None,
		'last_virtual_user_count':None,
		'total_global_count_failure':0,
		'total_global_downloaded_bytes':97876,
		'total_global_downloaded_bytes_per_second':8962.183,
		'total_iteration_count_failure':0,
		'total_iteration_count_success':10,
		'total_request_count_failure':0,
		'total_request_count_per_second':1.0988004,
		'total_request_count_success':12,
		'total_request_duration_average':495.75,
		'total_transaction_count_failure':0,
		'total_transaction_count_per_second':1.0072337,
		'total_transaction_count_success':11,
		'total_transaction_duration_average':496.9091
	}

TODO: This summary will also include SLA summaries in Jan 2020

Test Modifications

Metadata on a test can be modified after the test is complete, such as name, description, and status. With the required --testid argument as well, these modification arguments can be used individually or in combination with each other.

neoload --testid [test id] --updatename 'Some new test name' --updatedesc 'Some new description' --updatestatus 'PASSED|FAILED'

Additionally, plus and minus (+/-) operators can be used to append or remove the text specified from the existing name or description data. This is useful in combination with hashtag keywords to flag a test as baseline, candidate, or with an issue tracking id.

neoload --testid [old baseline test id] --updatedesc '-#baseline'
neoload --testid [new test id] --updatedesc '+#baseline'

Test Result listings and queries

TODO: will be coming in Jan 2020

Continuous Testing Examples

The main goal of the NeoLoad-CLI is to standardize the semantics of how load tests are executed across development, non-prod, and production environments. While the above instructions could be run from a contributor workstation, they can easily translated to various continuous build and deployment orchestration environments, as exampled:

• Jenkins
• Azure DevOps
• Gitlab
• Sorry AWS CodeBuild, haven't seen any F100 clients using the pform
• CircleCI, TBD when @punkdata gets back to @paulsbruce :)

Packaging the CLI with Build Agents

Many of the above CI examples include a step to explicitly install the NeoLoad CLI as part of the build steps. However, if you want the CLI baked into some build agent directly so that it is ready for use during a job, here's a Docker example:

For Docker builds:

• See the test harness Alpine-based Dockerfile

IDE Integrations

Since most of what we do in an IDE is create/edit code, we're mostly interested in how to:

• make it easy to write API tests in YAML (automatic syntax validation)
• validate that tests do not contain unanticipated errors even at small scale
• dry-run small (smoke) load tests locally so that code check-ins will work in CI/pipeline tests

Since the latter two cases are already covered by command-line semantics, our primary focus is to accelerate test authoring by providing NeoLoad as-code DSL (Domain-specific Language) validation and in some cases editor auto-complete.

Status of IDE / editor integrations

IDE / Editor	Syntax checks	Auto-complete	Setup steps
Visual Studio Code	[x]	[x]	see instructions

Contributing

Feel free to fork this repo, make changes, test locally, and create a pull request. As part of your testing, you should run the built-in test suite with the following command:

python3 -m pytest -v tests

*NOTE: omitting the --skipslow and --skipslas arguments also runs Docker-related attaching tests, which you will need to set environment variables up for in order to successfully run the test suite. An example of these variables can be found in example.bash_profile and can be addapted for Windows execution as well.