Microservice kinda implies usage of a container for me. How else would we google-scale it to serve all requests in parallel?
But thinking about, we probably have to use some more microservices, we can't put all that burden on the requester. So a dedicated service for compiling and executing in sandboxes would be necessary. Also, some local load balancers to control the flow and filter out the useless answers. So I'm not an expert on that devops-magic, but I guess this means ~12.5 billion pods fast enough result. Do Amazon or Google offer planetary scale for services?
alt Hacker News
Microservice kinda implies usage of a container for me. How else would we google-scale it to serve all requests in parallel?
But thinking about, we probably have to use some more microservices, we can't put all that burden on the requester. So a dedicated service for compiling and executing in sandboxes would be necessary. Also, some local load balancers to control the flow and filter out the useless answers. So I'm not an expert on that devops-magic, but I guess this means ~12.5 billion pods fast enough result. Do Amazon or Google offer planetary scale for services?