Debugging & Profiling Spinnaker Applications Live!
summary
Clones the clouddriver repo, finds the exact commit via hal version --bom, and adds JVM debug flags so IntelliJ remote debug connects over the service port plus 100; forwards kubectl port-forward 47102 and 7002 to hit the pod and trigger breakpoints. Diagnoses a credentials endpoint that degrades from ~3–6s to ~90s at scale. For profiling builds a custom Docker image that copies the Glowroot jar and libs, adds the Java agent, swaps the image, and exposes Glowroot agent on port 4000 to collect traces and JVM metrics.
AI-written summary · about
chapters (10)
- 0:00Start
- 5:12connect the debugger to this particular instance
- 10:05update our command-line argument with the appropriate port
- 12:03create a claw driver dot yml
- 12:36use halyard to redeploy
- 16:41drill down into the java source code
- 17:46forward the port from our local machine to the kubernetes pot
- 20:53set conditional breakpoints
- 26:13set the service settings for glow root
- 27:28add in the jvm options
related talks
25mDebugging at Scale in Production with kubectl debug
46mProfiling your Java Application A Beginner’s Guide
34mChaos Debugging: Finding and Fixing Microservice Weak Points
34mDebugging at Scale in Production - Deep into Your Containers with Kubectl Debug
24mDebugging at Scale in Production - Deep into your Containers
34mLive Production Debugging with Cloud Foundry and Stackdriver in 3 Easy Steps
30mCloud Native Debugging in Production - Dig Deep into your agents
52mPerfecting reliable code delivery for the cloud
