Infrastructure Scaling Strategies
This cluster discusses strategies for scaling web applications and infrastructure to handle sudden traffic spikes, such as those from Hacker News frontpage features, including autoscaling, vertical/horizontal scaling, cloud services like AWS, and challenges with databases and costs.
Activity Over Time
Top Contributors
Keywords
Sample Comments
It does not do automatic scaling, but, if you're expecting a one time spike you can just go with a huge server for that hour.
What? looks like you don't really understand the power of aws/serverless/Kubernetes/kafka/cassandra/graphql/react. when your app becomes instant hit after featuring in the frontpage of HN, you will have 1 billion daily active users and your business goes into flames because of new workload.
Bandwidth would be the concern, I think. They won't auto-scale machines up/down for you.
Sounds more expensive than scaling up more servers on demand.
This is "not that different" from getting a very high load spike do you guys not have some autoscaling setup?
There was a capacity crunch on the AWS nodes they rely on. I expect them to have a quality dial to pull back before overload. It would scale based on usage and server availablity.
Does your number of instances scale up and down frequently, or does it monotonically increase with growth?
At what point would you expect to need to scale out?
how..... does a web application scale infra to these mind boggling numbers instantly?
Using AWS or similar, you can upgrade to a huge server for just a few hours, then downgrade it once the traffic tapers.