Finding zombies in our systems: A real-world story of CPU bottlenecks

I know this is a hopelessly naive question but what is Pinterest even doing with all this AI training?

Interesting story.

I don't know anything about the Linux scheduler, but if one core out of 96 is busy chasing after those ghost allocations, why would a network thread not use one of the other 95 cores, and instead be stuck for seconds? Unless that one core is blocking the entire kernel, but then that would be a lot more visible instantly.

Jeez. Surely the ECS agent could just query for permission to join a cluster, and disable if disallowed?