Just spent 2 days panicking over running our workloads on #AKS #k8s using Windows Node Pools and a burst of CPU load would bring down the entire node despite setting CPU limits.

Turns out this was fixed in 1.18 and is beta right now, so I'm going to be shipping beta to prod AMA
If anyone has experience on HOW exactly the limits are gonna impact .NET Framework 4.8 apps... TELL ME WHAT YOU KNOW.

Does it gently respect them or does AKS have to put the container in timeout.
Y'all its STILL so much better than the alternative of deploying a single pod/container on 500+ dual core nodes on a separate cluster, so IDGAF LMAO
I swear 90% of my value prop for a company is that I find relevant github issues and PRs for the issues we face. Its the modern stackoverflow skillset.
YEAH! limiting to .5 (500m) CPU perfectly as a throttle, not an eviction.
You can follow @aidapsibr.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: