Optimizing for throughput
We are building a service fronted by Workers, which needs to see as high a throughput as possible. We are serving data blobs in about the ~24mb-64mb range, so they are medium sized.
When working through Workers using the default
fetch mechanism, things are fast but not as fast as they are at our endpoint. Is there anything we can do to optimize this?1 Reply
How is your request flow? Do you await the fetch and then respond to the client or stream the data as it comes in?