Hello!
On Wed, Nov 20, 2013 at 12:50:00AM -0800, John Watson wrote:
> This really only affects SSL connections, where if the buffer with
> chunk terminator is copied into the SSL buffer by itself. Since the
> new SSL buffer has not reached 16k, it's not flushed thus causing a
> delay in the client seeing it.
I don't think that unconditionally setting flush flag is a good
idea, it will cause unneeded work if a flush isn't needed.
It may make sense to set it (or move?) if previous buffer has it
set though.
--
Maxim Dounin
http://nginx.org/en/donation.html
_______________________________________________
nginx mailing list
nginx@nginx.org
http://mailman.nginx.org/mailman/listinfo/nginx
On Wed, Nov 20, 2013 at 12:50:00AM -0800, John Watson wrote:
> This really only affects SSL connections, where if the buffer with
> chunk terminator is copied into the SSL buffer by itself. Since the
> new SSL buffer has not reached 16k, it's not flushed thus causing a
> delay in the client seeing it.
I don't think that unconditionally setting flush flag is a good
idea, it will cause unneeded work if a flush isn't needed.
It may make sense to set it (or move?) if previous buffer has it
set though.
--
Maxim Dounin
http://nginx.org/en/donation.html
_______________________________________________
nginx mailing list
nginx@nginx.org
http://mailman.nginx.org/mailman/listinfo/nginx