On Aug 12, 2010, at 10:00 AM, mdipierro wrote: > For the page I am considering the fail requests are not a real failure > but declare a content-length of 19383 (wrong) instead of 19384 > (correct). Let's continue investigate...
Once the chunked encoding question is answered, it seems to me that we ought to create some pages (in examples?) expressly for the purpose of benchmarking. Make sure they're the same length, so we don't get the bogus errors, and create some variations: with and without session.forget, with and without a database access, etc.
