3 ms·
Yes, agreed. You're creating a resource to do the orchestration built with particular rules, rather than creating and applying those rules directly. I'm not sur
by nailer 7y ago
Yes, agreed. You're creating a resource to do the orchestration built with particular rules, rather than creating and applying those rules directly. I'm not sure the distinction is significant though - ultimately the rules in CF determine your scale-up/scale-down behaviour, directly or indirectly.
Perhaps more to the point of the other poster: autoscale groups existed pre-containers and microVMs exist after, orchestration being orthogonal to containment tech.
- scarface74 7y agoSort of playing devils advocate. But lambdas are containers. Instead of a Docker container, you have a zip file. Now, you can even have up to 5 layers in your lambda and it works even more like a building a Docker container.
- nailer 7y ago> But lambdas are containers. Nope. Technically they're MicroVMs, based on KVM not LXC. They're an isolation tech but containers aren't the only isolation tech (which is really the whole point of this discussion). https://firecracker-microvm.github.io/ https://firecracker-microvm.github.io/ Firecracker runs on Intel CPUs. 2 layers. No probs with you playing devil's advocate, you're a lot more polite than the other guy (who seems to have quietened down since pre-and-post LXC orchestration was pointed out to him).