Hi,
at work we tried to configure the Apache HTTP server to act as a reverse proxy with caching. We couldn't get it to work in more than three hours. At home I wrote exactly this in two hours: A reverse proxy that caches the requests: https://github.com/hauke96/tiny-http-proxy
It would be great if I get some feedback to the code: What could go wrong using it? What could be done more efficient/readable/elegant/shorter/...?
Regards Hauke
评论：
tsdtsdtsd:
tv64738:I believe you should use a custom http.Client instead of a direct http.Get, so you can set a meaningful timeout. And maybe you want to handle redirects in a custom way too. This is a nice article about that stuff. EDIT: also check the comments in that article, some good stuff there too.
theo_retiker:He should definitely plumb up the incoming context to the outgoing one, to get request cancellation.
-amir:Thanks for this article, I thought all the time that there's a default timeout like 30 seconds. But now I'm using the http.Client and http.Server (as mentioned in the comments, the default server has timeout issues too).
theo_retiker:Seems like the cache is not thread safe...
dahlma:Oh, yes, that's right and something I'll work on later. Thank for the hint :)
This is cool! I've written a proxy as well, unfortunately not something I can share but so far it's worked out pretty well. I'm storing the data in a DB instead of file system since I'm running the app on heroku, I need it to not rely on the filesystem (s3 may be an option but would need to look into latency)
I'm wondering though, when you save the information to the filesystem, does that retain headers?
