Anyway, I’ll only be able to get around to it when I have time, especially if the reproduction is complicated or expensive. Optimizations aren’t a particularly high priority at this stage, since they’re something that anyone could do in the meantime, and I have to get the 2.0 release ready.
Thanks! Anyone is welcome to help investigate further at this point.
No problem then… All I did is create a Django view that returns a static string and put it under gunicorn and put caddy infront of it.
My first tests were with curl. I decided to use ab to make testing higher request counts easier. I realized it was happening under no load so I used it for demonstration.
It takes the Django view 84ms or so from those logs. And the reverse proxy is taking 4 seconds. I figured it was a configuration issue. If it isn’t, so be it. I didn’t mean to bother you…
Certainly looking forward to the release but I can’t imagine I’ve done anything unusual.
Vpc is just the term Amazon uses for the space they create for you to run their computers in. I can’t imagine this demo costing more than a couple of dollars if you spend some time with it.