Closed reykjavikingur closed 8 years ago
On further research, the truncation seems to be caused by time-outs in very large non-HTML files from the proxy target. Why is harmon parses non-HTML files like images and JS files?? How do I restrict it to HTML only?
Never mind. I'm getting too many issues mixed up.
I'm running a proxy whose target uses gzip
content-encoding
andcontent-length
headers. The proxy works fine until I introduce the harmon middleware without even any selectors:app.use(harmon([], []))
Many of the proxy responses are getting truncated. I presume this is because the content is being decompressed (gunzip'd) but the content length is not being changed.
I've used harmon on proxy targets without gzip encoding or content lengths and have not had any issues. What do you recommend for a workaround?