I'm looking for advice on how to effectively capture and store very large memory crash dumps (over 100GB) from a Windows pod in Azure Kubernetes Service (AKS) after it crashes. It's essential that these dumps are saved without any corruption so that they can be extracted or reviewed later. The cluster is running on AKS, and I've previously attempted using a premium Azure disk (az-disk), but it hasn't been reliable in this scenario. I'm also considering the use of emptyDir but haven't experimented with that yet. Any suggestions or experiences would be really helpful, thanks!
3 Answers
It sounds like capturing large dumps from Windows pods might require a different approach. You should definitely look into restructuring your application to mitigate those memory dumps. Sometimes it’s a sign of something that might need tuning in your code.
I'd be interested to know if you find a solution! Your title threw me for a second—I was thinking of something totally different! Anyway, good luck!
Exactly! I'm just here for the laughs while figuring out the serious stuff.
What's your end goal here? Are you debugging an application? Understanding what you’re trying to achieve can help in advising better approaches. Also, I've struggled with Windows containers in Kubernetes myself, so I get the frustration.
Yes, I'm trying to debug the application, which is just a Windows pod creating those large memory dumps. Honestly, I wish I could avoid Windows containers altogether!
Haha, right? I definitely had a giggle at that, too. Maybe adding 'memory' to the title would help!