[nginx] internal location keepalive_requests issue

[baidu]

location /hls { 
error_page 404 = @hls; 
keepalive_requests 1000; 
} 

location @hls { 
# Serve HLS fragments 
types { 
application/vnd.apple.mpegurl m3u8; 
video/mp2t ts; 
} 
root /tmp; 
add_header Cache-Control no-cache; 
keepalive_requests 1000; 
} 

keepalive_requests must be large enough in this two location meanwhile 
if set keepalive_requests to 0 or 1 in /hls, keepalive_requests would not work in @hls

if (r->keepalive) { 
if (clcf->keepalive_timeout == 0) { 
r->keepalive = 0; 

} else if (r->connection->requests >= clcf->keepalive_requests) { 
r->keepalive = 0; 

} else if (r->headers_in.msie6 
&& r->method == NGX_HTTP_POST 
&& (clcf->keepalive_disable 
& NGX_HTTP_KEEPALIVE_DISABLE_MSIE6)) 
{ 
/* 
* MSIE may wait for some time if an response for 
* a POST request was sent over a keepalive connection 
*/ 
r->keepalive = 0; 

} else if (r->headers_in.safari 
&& (clcf->keepalive_disable 
& NGX_HTTP_KEEPALIVE_DISABLE_SAFARI)) 
{ 
/* 
* Safari may send a POST request to a closed keepalive 
* connection and may stall for some time, see 
* https://bugs.webkit.org/show_bug.cgi?id=5760 
*/ 
r->keepalive = 0; 
} 
} 

Note r->keepalive only effect that Connection filed in response header and set keepalive timer when finalize request. 

Why place this code block in ngx_http_update_location_config? Would be better place it where set keepalive timer? 

And sometimes nginx would send "Connection: keep-alive", but close connection in 1.10.1 version, and this would be nothing seriously

So why is it the expected behaviour?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nginx.org/pipermail/nginx-devel/attachments/20160827/2d163b30/attachment.html>