Previously, mruby context is wrongly shared by multiple patterns if
the underlying SharedDownstreamAddr is shared by multiple
DownstreamAddrGroups. This commit fixes it.
Previously, for Via, X-Forwarded-For, and X-Forwarded-Proto header
field, nghttpx only forwarded the last header field of each. With
this commit, nghttpx forwards all of them if it is configured to do
so.
This commit removes frontend-tls parameter, and adds
redirect-if-not-tls parameter parameter to --backend option. nghttpx
now responds to the request with 308 status code to redirect the
request to https URI if frontend connection is not TLS encrypted, and
redirect-if-no-tls parameter is used in --backend option. The port
number in Location header field is 443 by default (thus omitted), but
it can be configurable using --redirect-https-port option.
Previously, we have dedicated read buffer for each frontend
connection. With this commit, the buffer spaces are only used when
needed, and pooled if they are not used. This reduces memory usage
for idle client connections.
Previously we wrongly handles stream per connection when h2 backend
failed or closed. If upstream is h2 or spdy, streams which are not
associated to the failed h2 backend are also handled, which is
unnecessary.
We added --frontend-http2-window-size,
--frontend-http2-connection-window-size, --backend-http2-window-size,
and --backend-http2-connection-window-size option to replace existing
*-bits options. The old options are not flexible because they only
specify number of bits. Now we can specify integer value, with
possible g, m, and k unit. The old options are still available for
backend compatibility, but are deprecated.
There is a discussion in httpbis mailing list which argues that
dependency tree is for client, and changing it in server side is not
what client expects.
https://lists.w3.org/Archives/Public/ietf-http-wg/2016JulSep/0416.html
Currently, we make pushed stream depend on the parent stream of
associated stream (that is main HTML in most of the cases), so that
associated stream and pushed stream become siblings. In this case, we
also observed that these resources complete each other to get its
parent weight. This means that the delivery of associated stream is
delayed by pushed streams.
So at this moment, it is not a good idea to change pushed stream
priority in a way we do currently.
For HTTP/2, read timer starts when there is no downstream, and timer
stops when there is at least one downstream. For HTTP/1, read timer
starts when request handling finished, and timer stops when request
handling starts.
Tatsuhiro Tsujikawa