---
title: "BaoKhang (@baokhang) | Hashnode"
description: "Java Spring/Angular Architect | AI-Augmented Software Engineering"
url: "https://hashnode.com/@baokhang"
---

@baokhang

# BaoKhang

@baokhang · Joined July 2026

### [Predictive Test Selection Without SaaS Lock-In: Storing TIA Indexes in GitHub Actions Cache or S3](https://baokhang.hashnode.dev/predictive-test-selection-without-saas-lock-in-storing-tia-indexes-in-github-actions-cache-or-s3?utm_source=hashnode&utm_medium=feed)

Hosted predictive test selection products solve a real problem, and they come with a procurement conversation: another vendor, another bill, and a security review about sending build and test metadata

2h ago · 10 min read [baokhang.hashnode.dev](https://baokhang.hashnode.dev/predictive-test-selection-without-saas-lock-in-storing-tia-indexes-in-github-actions-cache-or-s3?utm_source=hashnode&utm_medium=feed)

### [How to Run Test Impact Analysis in Maven Without Breaking JaCoCo Code Coverage](https://baokhang.hashnode.dev/maven-test-impact-analysis-without-breaking-jacoco?utm_source=hashnode&utm_medium=feed)

Test impact analysis (TIA) and code coverage both want the same thing from your build: a seat on the test JVM's command line. Coverage tools attach a Java agent through argLine. Many TIA tools do too.

2h ago · 8 min read [baokhang.hashnode.dev](https://baokhang.hashnode.dev/maven-test-impact-analysis-without-breaking-jacoco?utm_source=hashnode&utm_medium=feed)

### [Why Maven Build Cache Still Re-Runs Your Entire Test Suite on One-Line Edits](https://baokhang.hashnode.dev/why-maven-build-cache-reruns-entire-test-suite?utm_source=hashnode&utm_medium=feed)

You install maven-build-cache-extension, your unchanged modules start restoring from cache in seconds, and PR builds feel fast. Then you edit one line in a utility class inside your biggest module and

3h ago · 8 min read [baokhang.hashnode.dev](https://baokhang.hashnode.dev/why-maven-build-cache-reruns-entire-test-suite?utm_source=hashnode&utm_medium=feed)

### [How We Cut Maven & JUnit 5 Test Runs by 61% Across Apache Repos Without Missing Regressions](https://baokhang.hashnode.dev/how-we-cut-maven-junit5-test-runs-61-percent-apache-repos?utm_source=hashnode&utm_medium=feed)

Most "predictive test selection" write-ups for Java come from vendors, and most of them give you a percentage with no way to check it. This one comes with the commit hashes. We replayed 800 consecutiv

8h ago · 9 min read [baokhang.hashnode.dev](https://baokhang.hashnode.dev/how-we-cut-maven-junit5-test-runs-61-percent-apache-repos?utm_source=hashnode&utm_medium=feed)

### [Why Kubernetes In-Place Memory Resize Gets Stuck "In Progress" on JVM Pods](https://baokhang.hashnode.dev/kubernetes-pod-resize-stuck-in-progress-jvm?utm_source=hashnode&utm_medium=feed)

You lowered a JVM pod's memory with an in-place resize, and the pod now shows PodResizeInProgress, an Error reason, and an old memory limit that never changes. The kubelet is not broken. It refuses to

17h ago · 10 min read [baokhang.hashnode.dev](https://baokhang.hashnode.dev/kubernetes-pod-resize-stuck-in-progress-jvm?utm_source=hashnode&utm_medium=feed)