Load TEsting Your App

CascadiaPHP 2019

Ian Littman / @iansltx

follow along at https://ian.im/loadcas19


QUestions We'll Answer

  • What types of tests exist, and when should I use them?
  • How can I match my load test with (anticipated) reality?
  • What does a real load test script look like on a small system?
  • How do I properly analyze results during and after my test?

Questions we won't answer

  • How do I use $otherPerfTestTool (!== 'k6')?
  • How can I set up clustered load testing?
  • How can I simulate far-end users?
  • How can I do deep application profiling? (Blackfire)
  • What about single-user load testing?

This is what we'll test with*


* More tools are listed at the end of this presentation.
** It's JS, but it uses goja, not V8 or Node, and doesn't have a global event loop yet.
*** I've used this on a project significantly more real than the example in this presentation, so that's a big reason we're looking at it today.

A Challengr Appears

This will be our system under test


  • Load Test vs. Stress Test
  • Soak Test vs. Spike Test
  • Smoke Test

Load TEst

  • <= expected peak traffic
  • Your system shouldn't break
  • If it does, it's a...

Stress Test

  • Increase traffic above peak || decrease available resources
  • Trying to break your system
  • Surfaces bottlenecks

Soak Test

  • Extended test duration
  • Watch behavior on ramp down
    as well as ramp up
  • Memory leaks
  • Disk space exhaustion (logs!)
  • Filled caches

Spike Test

  • Stress test with quick ramp-up
  • Woot.com at midnight
  • TV ad "go online"
  • System comes back online after downtime
  • Everyone hits your API via on-the-hour cron jobs

Smoke test

  • An initial test to confirm the system operates properly without a large amount of generated load
  • Do this before you load test
  • Pick your implementation...
    • Integration tests in your existing test suite
    • Load test script, turned down to one (thorough) iteration and one Virtual User (VU)

When should you run a load test?

  • When your application performance may change
    • Adding or removing features
    • Refactoring
    • Infrastructure changes
  • When your load profile may change
    • Initial app launch
    • Feature launch
    • Marketing pushes and promotions

How should I test?

How should I test?


What should I test?

  • Flows, not just single endpoints
  • Frequently used
  • Performance intensive
  • Business critical

Concurrent Requests != Concurrent Users

  • Think Time
  • API client concurrency
  • Caching (client-side or otherwise)

How not to model think time

Other Oversimplifications to avoid

  • No starting data in database
  • No parameterization
  • No abandonment at each step in the process
  • No input errors

Vary Your Testing

  • High-load Case: more expensive endpoints get called more often
  • Anticipated Case
  • Low-load Case: validation failures + think time


Auth + Fixture Data

Probabilities and input specs

THink times

Trends: How long did it take?

Now that our setup is done...

Let's make an HTTP Call

Make sure we fail successfully

Make sure we succeed successfully

Making simultaneous requests

checking simultaneous responses

Timing simultaneous responses

What's next?

let's create a challenge

Challenge Accepted?

Understand your load test tool

Aggregate your metrics repsonsibly

  • Average
  • Median (~50th percentile)
  • 90th, 95th, 99th percentile
  • Standard Deviation
  • Distribution of results
  • Explain (don't discard) your outliers

Keep it real

  • Use logs and analytics to determine your usage patterns
  • Run your APM (e.g. New Relic, Tideways) on your system under test
    • Better profiling info
    • Same performance drop as instrumenting production
  • Is your infrastructure code? (e.g. Terraform, CloudFormation)
    • Easier to copy environments
    • Cheaper to set up an environment for an hour to run a load test
  • Decide whether testing from near your env is accurate enough
  • Test autoscaling and/or load-shedding facilities

Warning: Tricky bottlenecks ahead

  • Just because a request is expensive
    doesn't mean it's the biggest source of load
  • As a system reaches capacity
    you'll see nonlinear performance degradation

Bottlenecks: Web Server + DAtabase

  • FPM workers/Apache processes
  • DB Connections
  • CPU + RAM utilization
  • Network utilization
  • Disk utilization (I/O or space)

Bottlenecks: Load Balancer

  • Network utilization/warmup
  • Connection count

Bottlenecks: External Services

  • Rate limits (natural or artificial)
  • Latency
  • Network egress

Bottlenecks: Queues

  • Per-job spin-up latency
  • Worker count
  • CPU + RAM utilization
    • Workers
    • Broker
  • Queue depth

Bottlenecks: Caches

  • Thundering herd
  • Churning due to cache evictions

Let's fix some bottlenecks...

Bonus material: More Tools

  • Tsung
    • Erlang (efficient, high volume from a single box)
    • Flexible (not just HTTP)
    • XML based config
  • The Grinder
    • Java-based
    • Java, Jython or Clojure scripts

BONUS MATERIAL: Even More Tools!

  • Artillery.io
    • Node-based
    • Simple stuff in Yaml, can switch to JS (including npm)
  • Molotov (by Mozilla)
    • Python 3.5+, uses async IO via coroutines
  • Locust
    • Python based
    • Can be run clustered
  • Wrk2
    • Built in C
    • Scriptable via Lua

What We Learned

  • What types of tests exist, and when you should use them
  • How to match load test with (anticipated) reality
  • What a real load test script looks like in K6
  • How to analyze results during and after your test

Further Reading

Thanks! Questions?