On Tue, Oct 20, 2015 at 12:15 AM, Andy Wang <awang@xxxxxxx> wrote: > > On 10/19/2015 05:08 PM, Yann Ylavic wrote: >> >> On Mon, Oct 19, 2015 at 11:44 PM, Andy Wang <awang@xxxxxxx> wrote: >>> >>> >>> Spdficially, note the 5 second delay between the first segment (No. 1854) >>> and the second data segment (1856). >> >> >> Does it correspond to the KeepAliveTimeout configured on the httpd side? >> (Note that if you did not configure it, the default value is also 5 >> seconds). >> >> Can you reproduce with "KeepAlive off"? >> > > Sorry, I forgot to mention that. Yeah disabling KeepAlive makes the problem > go away. Increasing the KeepAliveTimeout also increases the delay but only > to a maximum of 20 seconds. > > I tried KeepAliveTimeout 30 and 60 and made a fresh connection and saw the > second data packet arrive after a 20 second delay. I forwarded the discussion on the dev@ mailing-list, since httpd is delaying the write on the socket to fill in the pipe. I think this should be limited to pipelined request, but would like to ear others on this. I will report here if you're not on dev@. Regards, Yann. --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscribe@xxxxxxxxxxxxxxxx For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx