I'm seeking advice on how to effectively capture and store very large full memory crash dumps (over 100GB) from a Windows pod in Azure Kubernetes Service (AKS) after a crash. I want to ensure that these dumps are saved without any corruption and can be easily downloaded or inspected later. For context, I've already tried using a premium Azure disk (az-disk), but it hasn't proven reliable for this situation. I'm also looking into using emptyDir but haven't experimented with that yet. Any thoughts would be greatly appreciated!
3 Answers
For Windows pods, handling large dumps often requires a different approach. If you're consistently getting large dumps, it might be worth revisiting your application's memory management. A smaller footprint could make a big difference, and sometimes refactoring is necessary.
What's your main goal with these dumps? Are you focusing on debugging specific issues? Knowing that could help tailor the advice better.
Yeah, I'm primarily trying to debug my application. It just generates these huge dumps, and honestly, I find working with Windows containers in Kubernetes quite frustrating.
It's definitely a tricky situation! Have you considered other storage options like Azure Blob Storage? That might be a more robust solution for handling large files. Also, using a sidecar container to facilitate the dump could help with stability.
Interesting thought! I'll check out Azure Blob Storage for this.
Thanks for the suggestion! I’ll look into the sidecar approach too.
I can relate! Windows containers can be a real headache in a Kubernetes setup.