https://marktaguiad.dev/tags/testinag/ Recent content in Testinag on marktaguiad.dev Hugo en-us marktaguiad@marktaguiad.dev (Mark Taguiad) marktaguiad@marktaguiad.dev (Mark Taguiad) marktaguiad.dev Fri, 03 Apr 2026 00:00:00 +0000 https://marktaguiad.dev/post/k8s-istio-quatro/ Fri, 03 Apr 2026 00:00:00 +0000marktaguiad@marktaguiad.dev (Mark Taguiad) https://marktaguiad.dev/post/k8s-istio-quatro/ <img class="theme-image" src="https://marktaguiad.dev/images/devops/k8s-notes/istio/ab-testing/k8s-ab-testing-001.png" data-light="/images/devops/k8s-notes/istio/ab-testing/k8s-ab-testing-001.png" data-dark="/images/devops/k8s-notes/istio/ab-testing/k8s-ab-testing-dark-001.png" alt="Architecture Diagram" > <p>A/B testing and Canary Deployment allows you to route traffic between different versions of your application to compare performance, behavior, or user experience.</p> <p>With Istio, you can control traffic without changing application code—just by configuring the service mesh.</p> <p>Replicate this using this <a href="https://github.com/mcbtaguiad/istio-demo">repo</a>.</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="ln">1</span><span class="cl">kubectl apply -k kube/ab-testing/demo </span></span></code></pre></div><h1 id="table-of-contents">Table of Contents</h1> <nav id="TableOfContents"> <ol> <li><a href="#ab-testing-vs-canary-deployment">A/B Testing vs Canary Deployment</a></li> <li><a href="#ab-testing">A/B Testing</a> <ol> <li><a href="#labels">Labels</a></li> <li><a href="#service">Service</a></li> <li><a href="#destination-rule">Destination Rule</a></li> <li><a href="#virtual-service">Virtual Service</a></li> <li><a href="#rollout-strategy">Rollout Strategy</a></li> <li><a href="#roll-back-strategy">Roll-back Strategy</a></li> </ol> </li> <li><a href="#canary-deployment">Canary Deployment</a> <ol> <li><a href="#rollout-strategy-1">Rollout Strategy</a></li> </ol> </li> <li><a href="#observe-metrics">Observe Metrics</a></li> </ol> </nav> <h3 id="ab-testing-vs-canary-deployment">A/B Testing vs Canary Deployment</h3> <table> <thead> <tr> <th>Aspect</th> <th>Canary Deployment</th> <th>A/B Testing</th> </tr> </thead> <tbody> <tr> <td>Purpose</td> <td>Safe rollout</td> <td>Experimentation</td> </tr> <tr> <td>Traffic split</td> <td>Gradual (10% → 100%)</td> <td>Fixed (often 50/50)</td> </tr> <tr> <td>Decision basis</td> <td>Errors, latency</td> <td>User behavior, metrics</td> </tr> <tr> <td>End goal</td> <td>Replace old version</td> <td>Pick best variant</td> </tr> <tr> <td>Versions</td> <td>Usually same feature</td> <td>Different UX/features</td> </tr> </tbody> </table> <h3 id="ab-testing">A/B Testing</h3> <h4 id="labels">Labels</h4> <p>Label your deployments properly.</p>