Client times out on (304) requests

[saravsars]

For 200 requests everything works fine

For 304 requests , it takes 1 minute to complete

In error log

What is the issue here??

OS: Ubuntu 12.04
Nginx version: 1.5.6
Pagespeed : 1.6.29.5

Thanks

[oschaaf]

@saravsars I fail to reproduce this with Ubuntu 12.04.3/nginx 1.5.6/Pagespeed 1.6.29.5.
Would you be able to crank up the logging level to debug, reproduce, and post the error log content?

I tried with chrome, but also with curl:

oschaaf@ubuntu:~$ curl -vv 'http://192.168.185.185/mod_pagespeed_example/images/xCuppa.png.pagespeed.ic.jf3PtqQ39N.png' 'http://192.168.185.185/mod_pagespeed_example/images/xCuppa.png.pagespeed.ic.jf3PtqQ39N.png' -H 'Pragma: no-cache' -H 'Accept-Encoding: gzip,deflate,sdch' -H 'Host: 192.168.185.185' -H 'Accept-Language: nl-NL,nl;q=0.8,en-US;q=0.6,en;q=0.4' -H 'User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_8_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/30.0.1599.69 Safari/537.36' -H 'Accept: image/webp,*/*;q=0.8' -H 'Cache-Control: max-age=0' -H 'If-None-Match: W/"0"' -H 'Connection: keep-alive' -H 'If-Modified-Since: Mon, 07 Oct 2013 08:04:06 GMT' -H 'Referer: http://192.168.185.185/mod_pagespeed_example/rewrite_css_images.html?PageSpeed=on&PageSpeedFilters=rewrite_css,rewrite_images,prioritize_critical_css' --compressed
* About to connect() to 192.168.185.185 port 80 (#0)
*   Trying 192.168.185.185... connected
> GET /mod_pagespeed_example/images/xCuppa.png.pagespeed.ic.jf3PtqQ39N.png HTTP/1.1
> Pragma: no-cache
> Accept-Encoding: gzip,deflate,sdch
> Host: 192.168.185.185
> Accept-Language: nl-NL,nl;q=0.8,en-US;q=0.6,en;q=0.4
> User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_8_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/30.0.1599.69 Safari/537.36
> Accept: image/webp,*/*;q=0.8
> Cache-Control: max-age=0
> If-None-Match: W/"0"
> Connection: keep-alive
> If-Modified-Since: Mon, 07 Oct 2013 08:04:06 GMT
> Referer: http://192.168.185.185/mod_pagespeed_example/rewrite_css_images.html?PageSpeed=on&PageSpeedFilters=rewrite_css,rewrite_images,prioritize_critical_css
> 
< HTTP/1.1 304 Not Modified
< Server: nginx/1.5.6
< Date: Mon, 07 Oct 2013 08:18:40 GMT
< Connection: keep-alive
< X-Page-Speed: 1.6.29.5-3346
< 
* Connection #0 to host 192.168.185.185 left intact
* Re-using existing connection! (#0) with host 192.168.185.185
* Connected to 192.168.185.185 (192.168.185.185) port 80 (#0)
> GET /mod_pagespeed_example/images/xCuppa.png.pagespeed.ic.jf3PtqQ39N.png HTTP/1.1
> Pragma: no-cache
> Accept-Encoding: gzip,deflate,sdch
> Host: 192.168.185.185
> Accept-Language: nl-NL,nl;q=0.8,en-US;q=0.6,en;q=0.4
> User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_8_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/30.0.1599.69 Safari/537.36
> Accept: image/webp,*/*;q=0.8
> Cache-Control: max-age=0
> If-None-Match: W/"0"
> Connection: keep-alive
> If-Modified-Since: Mon, 07 Oct 2013 08:04:06 GMT
> Referer: http://192.168.185.185/mod_pagespeed_example/rewrite_css_images.html?PageSpeed=on&PageSpeedFilters=rewrite_css,rewrite_images,prioritize_critical_css
> 
< HTTP/1.1 304 Not Modified
< Server: nginx/1.5.6
< Date: Mon, 07 Oct 2013 08:18:40 GMT
< Connection: keep-alive
< X-Page-Speed: 1.6.29.5-3346
< 
* Connection #0 to host 192.168.185.185 left intact
* Closing connection #0
oschaaf@ubuntu:~$ 

[saravsars]

@oschaaf error log content for 304 request : http://pastebin.com/HgXJB4Qk

Can you share your nginx configuration file ??

Thanks.

[oschaaf]

@saravsars thanks for the log, I'll have a look
My nginx configuration: http://pastebin.com/CJrwEMz9

[oschaaf]

@saravsars I would also be interested in your nginx.conf and ./configure.
Could ngx_pagespeed be doing something here that doesn't work well with another nginx module?

[saravsars]

@oschaaf I'm using your configuration now but the issue is not resolved.

nginx.conf : http://pastebin.com/xz8wrhYF

nginx configure :

nginx version: nginx/1.5.6
configure arguments: --prefix=/opt/nginx --user=nginx --group=nginx --add-module=/home/Downloads/ngx_pagespeed-master --add-module=/home/Downloads/ngx_cache_purge-master --with-debug

[saravsars]

@oschaaf strangely this configuration and your configuration works if I enable defer-javascript filter. When I disable defer-javascript client timed out error occurs.

Any reason for it ??

nginx.conf

worker_processes 1;

events {
worker_connections 1024;
}

http {

pagespeed on;

server {
listen 9071;
server_name localhost;
pagespeed FileCachePath "/opt/ngx_pagespeed/";
pagespeed RewriteLevel CoreFilters;
pagespeed EnableFilters defer_javascript;

location /
{
    root html/banner;
    index index.html;
}   

}

sendfile on;

keepalive_timeout 65;

set up gzip

gzip on;
gzip_vary on;
gzip_http_version 1.0;

default_type application/octet-stream;

}


[oschaaf]

The ./configure and nginx.conf you are using are really similar to mine.
One thing I noted in the error log you posted was this:

2013/10/07 14:02:35 [debug] 9340#0: *36 ps fetch handler: /index.html
2013/10/07 14:02:35 [debug] 9340#0: *36 HTTP/1.1 200 OK
Server: nginx/1.4.2

Did you perhaps switch testing to 1.4.2?

If defer_javascript makes the issue go away, the only thing I can come up with, is that you are hitting a corner case with either the timings or file sizes involved.
I suspect that disabling keepalives will resolve the problem, but I don't think that is a proper workaround.

[1] http://wiki.nginx.org/HttpCoreModule#keepalive_disable

[saravsars]

@oschaaf yeah I'm testing both 1.4.2 and 1.5.6.

Setting keepalive_timeout 0; solves the issue but I need some other solution.

Thanks for quick response 👍

[oschaaf]

@saravsars
Would you be able to post your website here? Or pm it if you don't want to post it public?
I feel it is like a long shot, but that way, we might be able to hit the same corner case and reproduce.

[saravsars]

@oschaaf I'll try to post it public asap.

BTW Im seeing this error many times in error.log

It is usual one ??
Full error.log : https://docs.google.com/file/d/0B9HIahYpqDmdNmUzNnNsbVF2cFE/edit?usp=sharing

Thanks

[oschaaf]

@saravsars Would you be able to try the current master?
@jeffkaufman just opened and diagnosed #538, and I think that might be related - which would mean this should be fixed on the current master.

[adegtyarev]

You may want to set keepalive_timeout 0; to only on pagespeed-specific location to workaround the problem:

location ~ "\.pagespeed\.([a-z]\.)?[a-z]{2}\.[^.]{10}\.[^.]+" {
  keepalive_timeout 0;
}

By the way, I only have these options related to pagespeed:

pagespeed on;
pagespeed FileCachePath /var/cache/ngx_pagespeed;

nginx: 1.4.3
pagespeed: 1.6.29.5-3346

[saravsars]

@oschaaf @degtyarevalexey yeah using keepalive_timeout 0; for pagespeed specific resources.

Thanks 👍

[tuurbo]

Having the same problem as saravsars. Using keepalive_timeout 0 fixed the problem for me as well, but is it safe to use keepalive_timeout 0 as described above without causing any problems?

ubuntu: 13.10
nginx: 1.4.3
pagespeed: 1.6.29.5

./configure --add-module=$HOME/ngx_pagespeed --with-http_ssl_module --with-openssl=/usr/bin/openssl

pagespeed on;
pagespeed FileCachePath /var/ngx_pagespeed_cache;

location ~ "\.pagespeed\.([a-z]\.)?[a-z]{2}\.[^.]{10}\.[^.]+" {
    add_header "" "";
    keepalive_timeout 0;
}
location ~ "^/ngx_pagespeed_static/" { }
location ~ "^/ngx_pagespeed_beacon$" { }
location /ngx_pagespeed_statistics { allow 127.0.0.1; deny all; }
location /ngx_pagespeed_message { allow 127.0.0.1; deny all; }
location /pagespeed_console { allow 127.0.0.1; deny all; }

Edit

For whatever reason putting keepalive_timeout 0 inside the pagespeed location regex stopped working for me. Has to be put outside of it.