On Tue, Apr 21, 2009 at 12:30:57PM -0500, AMP Admin wrote: > I see a lot of the following in our log. Anyone know what might be causing this? > > 15399#0: *15062 client *.*.*.* closed keepalive connection (104: Connection reset by peer) MSIE has closed a keepalive connection (this is normal). > 15401#0: *15199 client *.*.*.* closed keepalive connection A browser has cby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 02:49:33PM -0400, Jonathan Vanasco wrote: > This command serves the right directories, and pipes them through php: > > location /careers) { > root /home/artwelove/_current/web/www.artwelove.com-news/news; > include /usr/local/nginx/_macros/php.conf; > index index.php index.html; > } > > > This command servesby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 12:41:16PM -0700, Payam Chychi wrote: > hey guys, > > what is the main difference between these two? reading the > documentation it looks like both set the ip address of the client on > the request which allows you to pass the real client ip address to a > backend server via nginx proxy... or i could be totally wrong here! > > I was lookby Igor Sysoev - Nginx Mailing List - English
On Mon, Apr 20, 2009 at 03:20:59PM -0700, Chris Cortese wrote: > Thanks for this release! I've been trying to get nginx going on Windows > for a while now. > > With this latest build though, all I can get is "No input file > specified". I thought maybe I was specifying paths in a way it didn't > like but I've tried every variation on the Windows paths I caby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 01:11:45PM -0700, Payam Chychi wrote: > Hey, > > I noticed that the graceful reload was not actually loading the new > config changes and so ive had to use a hard down/start I never saw this. Anyway, nginx should write the reason in error_log. > kill -quit $master works > > i did also notice that nginx will not start with the following iby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 04:43:05PM +0200, Roger Pack wrote: > > Nginx server in front could take care of. I'd rather see something like > > Libevent (i.e. an asynchronous HTTP server engine) with embedded Python. > > The core of Nginx is probably similar to Libevent with its HTTP layer. > > Would be interesting to find out how Nginx core compares to Libevent >by Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 04:27:24PM +0100, Istv?n wrote: > 2009/4/21 Igor Sysoev <is@rambler-co.ru> > > > On Tue, Apr 21, 2009 at 04:52:09PM +0200, Roger Pack wrote: > > > > > >> I realized that during performance testing, all what I can tell you: > > use the > > > >> httperf instead if you have a chance. > > > >>by Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 05:36:01PM +0200, Roger Pack wrote: > > > I meant that nginx is built with FD_SETSIZE=1024 > > What does nginx do when it receives more connections than 1024? Does it > try to override the fdset instance and possibly tromp on other memory > (I'm not sure if that's a concern or not, even)? > > does it reject incoming connections wheby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 09:50:09AM -0700, Payam Chychi wrote: > hey Guys, > > Anyone else notice that in the resent *.52 release stopping/starting > nginx does not restart the process? I manually have to kill -9 the > master and each child... BTW, why do you use stop/start instead of graceful reload ? -- Igor Sysoev http://sysoev.ru/en/by Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 05:43:32PM +0000, Jim Ohlstein wrote: > I installed 0.7.52 on several Virtuozzo containers. All are running CentOS, some 32 bit and some 64. After reading your post I checked them all. One (and only one) is having a similar issue. It's a 64 bit container. All of the others are fine and all use an identical init script. Do you able to stop/reload nginx by head:by Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 04:52:09PM +0200, Roger Pack wrote: > >> I realized that during performance testing, all what I can tell you: use the > >> httperf instead if you have a chance. > >> (do not forget to modify the FD_SETSIZE to higher value) > > > > BTW, nginx/Windows is currently build with FD_SETSIZE. > > did you mean "without defby Igor Sysoev - Nginx Mailing List - English
On Sun, Apr 19, 2009 at 09:48:54AM -0700, Michael Shadle wrote: > make sure nginx has the "realip" module (nginx -V, look for > --with-http_realip_module) - you could try it and run nginx -t to > check the conf file to see if it's in your compile too. > > Double check the header isn't a different case like > X-FB-User-Remote-Addr or something (if you saw this iby Igor Sysoev - Nginx Mailing List - English
On Sun, Apr 19, 2009 at 01:43:45PM -0700, rkmr.em@gmail.com wrote: > Hi Michael, > Thanks a lot for this. > > I want to do this only if this header is available. IF it is not available, > it should the client ip should be the default one. > how can i do that? If the header is not available client address is not changed. > thanks again > > On Sun, Apr 19by Igor Sysoev - Nginx Mailing List - English
On Sun, Apr 19, 2009 at 09:48:54AM -0700, Michael Shadle wrote: > make sure nginx has the "realip" module (nginx -V, look for > --with-http_realip_module) - you could try it and run nginx -t to > check the conf file to see if it's in your compile too. > > Double check the header isn't a different case like > X-FB-User-Remote-Addr or something (if you saw this iby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 10:52:53AM +0100, Istv?n wrote: > Hi! > > Please note that ab itself a huge bottleneck and it gives you zero > information about nginx performance. As you can see you hit the 100% CPU > -what comes from ab load-. Yes, running ab on the same host is benchmark both ab and nginx. > I realized that during performance testing, all what I can tellby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 05:43:48PM +0800, Weibin Yao wrote: > > Just note for everyone: "-c 10" (concurrency 10) is far from real world > > usage. The select() method used by nginx is poor scaleable for concurrency > > more than 100-200. > It seems that the most suitable concurrency for my PC is 1-200. > > More tests in test case3 for Native Nginx inby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 07:38:36AM +0200, John Mccarthy wrote: > AMP Admin wrote: > > Prolly shouldn't run this as root. > > > > My guess would be to change user root root; to user nginx nginx; and see > > what happens. > > Thanks for the response. It seems the problem was that on my Mac there > was no group 'root', so I put 'wheel' and nginx starby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 01:20:05PM +0700, Glen Lumanau wrote: > I still wonder which one is better for performance > > Pure nginx + php-fpm > > Or > > Nginx + apache2 + proxy_cache > > Anybody knows about this? Can not say about apache2 vs php-fpm, but you can use fastcgi_cache too. -- Igor Sysoev http://sysoev.ru/en/by Igor Sysoev - Nginx Mailing List - English
On Sat, Apr 18, 2009 at 08:27:11PM +0200, Joe Bofh wrote: > Igor, > > It looks like nginx now has the beginnings of caching which is exciting > as I am looking forward to comparing this with varnish. > > I did notice in the docs that the key/filename is a md5 hash. Have you > looked at the recent MurmurHash/FNV-1a hashes? > > They are at least 2x faster thaby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 03:51:57PM +0800, Weibin Yao wrote: > >BTW, could someone run the simple small static file benchmark for > >nginx/Windows, nginx/Cygwin, Apache2 and IIS (from another host): > > >ab -n 1000 -c 1 ... > >ab -n 1000 -c 10 ... > >ab -n 1000 -c 1 -k ... " > > This is my rough test results. It seems that the > Windows vby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 03:10:31PM +0700, Glen Lumanau wrote: > Is there any help page regarding fastcgi_cache? Or it's the same format as > proxy_cache? Yes, just s/proxy/fastcgi/. The single difference is that fastcgi cache requires fastcgi_cache_key localhost:9000$request_uri; while proxy_cache may use default one, similar to this: proxy_cache_key $scheme$proxyby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 03:51:57PM +0800, Weibin Yao wrote: > >BTW, could someone run the simple small static file benchmark for > >nginx/Windows, nginx/Cygwin, Apache2 and IIS (from another host): > > >ab -n 1000 -c 1 ... > >ab -n 1000 -c 10 ... > >ab -n 1000 -c 1 -k ... " > > This is my rough test results. It seems that the > Windows vby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 03:44:28PM +0700, Glen Lumanau wrote: > I already tried this > > Here's the config > > fastcgi_cache_path /data/nginx/cache levels=1:2 keys_zone=one:10m > inactive=7d max_size=200m; > fastcgi_temp_path /data/nginx/temp; > > location @joomla { > fastcgi_pass 127.0.0.1:9000; > fastby Igor Sysoev - Nginx Mailing List - English
On Tue, Apr 21, 2009 at 08:06:04AM +0200, Joe Bofh wrote: > Understanding that cpu is not a primary factor (which is a good thing), > it still does not answer the question I asked. > > The fact is some apps perform better in 32bit mode and others in 64bit. > I just wanted to see if anyone had done testing in this area. I ran some time ago 32-bit nginx on FreeBSD/amd64 6by Igor Sysoev - Nginx Mailing List - English
On Mon, Apr 20, 2009 at 04:42:12PM +0200, Joaquin Cuenca Abela wrote: > 2009/4/20 ????????? <www@lc365.net>: > >> Changes with nginx 0.7.52 20 Apr 2009 > >> > >> *) Feature: the first native Windows binary release. > >> > > > > A really good news for me. I'm waiting for it for a long time. > > > > I think many peopleby Igor Sysoev - Nginx Mailing List - English
On Mon, Apr 20, 2009 at 03:13:33PM +0400, Igor Sysoev wrote: > On Mon, Apr 20, 2009 at 06:38:52PM +0800, Weibin Yao wrote: > > > Thanks for your hardwork. Do you have a plan to release the source of Windows version? > > Yes, after I will adopt ./configure for MSYS. > Currently I ./configure nginx/Windows on FreeBSD using Wine in some places > and I do want to suby Igor Sysoev - Nginx Mailing List - English
On Mon, Apr 20, 2009 at 06:38:52PM +0800, Weibin Yao wrote: > Thanks for your hardwork. Do you have a plan to release the source of Windows version? Yes, after I will adopt ./configure for MSYS. Currently I ./configure nginx/Windows on FreeBSD using Wine in some places and I do want to support this procedure. -- Igor Sysoev http://sysoev.ru/en/by Igor Sysoev - Nginx Mailing List - English
On Sun, Apr 19, 2009 at 09:12:37PM -1000, Kevin Worthington wrote: > Is there a special way to start the windows version? Double-clicking > did nothing, and running from a prompt did nothing. cd c:\ unzip nginx-0.7.52.zip ren nginx-0.7.52 nginx cd nginx start nginx Control: nginx -s [ stop | quit | reopen | reloadf ] For problems look in c:\nginx\logs\error.log or in Evenby Igor Sysoev - Nginx Mailing List - English
Changes with nginx 0.7.52 20 Apr 2009 *) Feature: the first native Windows binary release. *) Bugfix: in processing HEAD method while caching. *) Bugfix: in processing the "If-Modified-Since", "If-Range", etc. client request header lines while caching. *) Bugfix: now the "Set-Cookie" and "by Igor Sysoev - Nginx Mailing List - English
On Wed, Apr 15, 2009 at 12:26:15PM -0700, Michael Nachbaur wrote: > On 15-Apr-09, at 11:17 AM, Igor Sysoev wrote: > > >On Wed, Apr 15, 2009 at 10:10:07AM -0700, Michael Nachbaur wrote: > > > >>I'm developing a web application that makes heavy use of nginx's XSLT > >>support (Thank you for that!) About half of the content is in static > >>XMLby Igor Sysoev - Nginx Mailing List - English