{"id":97,"date":"2009-12-16T01:10:13","date_gmt":"2009-12-16T05:10:13","guid":{"rendered":"http:\/\/blog.litespeedtech.com\/?p=97"},"modified":"2009-12-16T01:10:13","modified_gmt":"2009-12-16T05:10:13","slug":"litespeed-lsapi-vs-apache-passenger-vs-nginx-passenger","status":"publish","type":"post","link":"https:\/\/lswp.store\/index.php\/2009\/12\/16\/litespeed-lsapi-vs-apache-passenger-vs-nginx-passenger\/","title":{"rendered":"LiteSpeed + LSAPI vs. Apache + Passenger vs. Nginx + Passenger"},"content":{"rendered":"

To perform a fair benchmark requiring a lot of preparation, as you need to know the detail of each product and set appropriate configuration. Recently, we got chance to do some internal testing. It\u2019s experimental, as all tests were run within 1 box. We bind LSWS to 1 CPU, and leave another CPU to ab. The result here is good enough to make some points and may be useful to our users, so we would like to share in a series of blog posts first. Feel free to provide any feedback, so we can improve in our formal benchmark test later on.<\/p>\n

In our last test, we studied the memory consumption of Ruby vs Ruby Enterprise Edition using Typo as a test case. Since Typo itself is a bottleneck when cache is disabled, performance among different web servers cannot be revealed.<\/p>\n

This naturally leads to our next interested topic \u201cComparing the performance of the interface between web server and Ruby engine<\/span><\/strong>\u201d.<\/p>\n

This time, we focus on performance of the interface between web server and Ruby engine by using the simplest \u201cHello World\u201d without database. We tested 3 configurations, all based on Ruby Enterprise 1.8.7: LiteSpeed\/4.1RC2 + REE, Apache\/2.2.14 (worker MPM) + REE + Phusion Passenger and Nginx\/0.7.61 \u00a0+ REE + Phusion Passenger.<\/p>\n

Test method is still based on\u00a0http:\/\/www.rubyenterpriseedition.com\/comparisons.html<\/a><\/p>\n

Software Version<\/span><\/strong>:<\/span><\/span> LSWS 4.1RC2, Nginx\/0.7.61, Apache\/2.2.14(Worker MPM),\u00a0Ruby Enterprise Edition 1.8.7, Rails 2.2, Passenger 2.2.5<\/span><\/span><\/p>\n

Test Result<\/span><\/strong><\/p>\n\n\n\n\n\n\n
Configuration<\/td>\nReq\/Sec<\/td>\nMemory Usage<\/td>\n<\/tr>\n
LiteSpeed + REE + LSAPI<\/td>\n280<\/td>\n146M ~ 158M<\/td>\n<\/tr>\n
Nginx + REE + Passenger<\/td>\n212<\/td>\n153M<\/td>\n<\/tr>\n
Apache(worker MPM) + REE + Passenger<\/td>\n200<\/td>\n178M<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Conclusion<\/span><\/strong><\/p>\n

Based on above simple tests, LiteSpeed + REE + LSAPI are the preferred configuration.<\/p>\n

<\/p>\n

\n

Details<\/strong><\/p>\n\n\n\n\n\n
Client Software<\/td>\nApacheBench (ab) 2.3<\/td>\n<\/tr>\n
Server Hardware<\/td>\nDual Athlon MP 2000+\/266FSB\/256KB L2 Cache
\nMemory: 1GB PC2100
\nMotherboard: MSI K7D
\nHard Drive: 36GB 1000RPM SCSI drive<\/td>\n<\/tr>\n
Server OS<\/td>\nLinux CentOS 5.4<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
This test focuses on performance, so the simplest \u201cHello World\u201d without database is used.<\/div>\n
LSWS 4.1RC2 – Bind listener to 1 processor, disable cache. (equivalent to 1cpu license)<\/div>\n

LiteSpeed + REE + LSAPI<\/strong><\/p>\n

LSWS 4.1RC2 + Ruby Enterprise 1.8.7 + Rails 2.2 + Hello
\n(Server–>Ruby Rails–>Ruby Path:\/usr\/local\/ruby-enterprise-1.8.7-2009.10\/bin\/ruby)
\nRails Context max connections: 8
\n~>(date;free -m;ab -n 20000 -c 100 http:\/\/127.0.0.1:48089\/hello;date;free -m) 2>&1|tee ttt;echo;egrep “(buffers\/cache)|(Requests per second)” ttt
\n<\/span><\/p><\/blockquote>\n\n\n\n\n\n\n\n
Round<\/td>\nFree Memory (M)
\nBefore<\/td>\n		Free Memory (M)
\nAfter<\/td>\n		Memory Usage
\n( Before – After)<\/td>\n		Request\/Second
\n[#\/sec](mean)<\/td>\n<\/tr>\n
1<\/td>\n876<\/td>\n730<\/td>\n146M<\/td>\n281.40<\/td>\n<\/tr>\n
2<\/td>\n877<\/td>\n734<\/td>\n143M<\/td>\n281.59<\/td>\n<\/tr>\n
3<\/td>\n877<\/td>\n720<\/td>\n157M<\/td>\n267.05<\/td>\n<\/tr>\n
4<\/td>\n877<\/td>\n719<\/td>\n158M<\/td>\n279.84<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Test again: restart LiteSpeed after each round of test<\/span><\/p><\/blockquote>\n\n\n\n\n\n\n\n
Round<\/td>\nFree Memory (M)
\nBefore<\/td>\n		Free Memory (M)
\nAfter<\/td>\n		Memory Usage
\n( Before – After)<\/td>\n		Request\/Second
\n[#\/sec](mean)<\/td>\n<\/tr>\n
1<\/td>\n873<\/td>\n772<\/td>\n101M<\/td>\n281.63<\/td>\n<\/tr>\n
2<\/td>\n872<\/td>\n747<\/td>\n125M<\/td>\n227.86<\/td>\n<\/tr>\n
3<\/td>\n872<\/td>\n714<\/td>\n158M<\/td>\n280.69<\/td>\n<\/tr>\n
4<\/td>\n872<\/td>\n714<\/td>\n158M<\/td>\n281.00<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Nginx + REE + Passenger <\/strong><\/p>\n


\nNginx\/0.7.61 + Ruby Enterprise 1.8.7 + Passenger 2.2.5 + Rails 2.2 + Hello
\ntest script:
\n~>(date;free -m;ab -n 20000 -c 100 127.0.0.1:48130\/hello;date;free -m)2>&1|tee ttt;echo;egrep “(buffers\/cache)|(Requests per second)” ttt
\nTo stop Nginx: \/usr\/local\/nginx-passenger\/sbin>.\/nginx -s stop
\nTo start Nginx: \/usr\/local\/nginx-passenger\/sbin>.\/nginx
\n<\/span><\/p><\/blockquote>\n

    Notes:<\/p>\n
  1. There are totally 8 ruby processes running during the test by checking “top” and “pstree”. That is why we set LiteSpeed rails context max connections to 8 to compare.<\/li>\n
  2. Nginx+REE+Passenger won’t free memory immediately after \u201cab\u201d testing, unlike LiteSpeed which frees memory immediately after \u201cab\u201d finishes. Therefore Nginx has to be restarted after each round of test.<\/li>\n<\/ol>\n\n\n\n\n\n\n
    Round<\/td>\nFree Memory (M)
    \nBefore<\/td>\n    		Free Memory (M)
    \nAfter<\/td>\n    		Memory Usage
    \n( Before – After)<\/td>\n    		Request\/Second
    \n[#\/sec](mean)<\/td>\n<\/tr>\n
    1<\/td>\n869<\/td>\n715<\/td>\n154M<\/td>\n211.47<\/td>\n<\/tr>\n
    2<\/td>\n868<\/td>\n715<\/td>\n153M<\/td>\n211.84<\/td>\n<\/tr>\n
    3<\/td>\n868<\/td>\n715<\/td>\n153M<\/td>\n213.39<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

    Apache + REE + Passenger <\/strong><\/p>\n


    \nApache\/2.2.14(Worker MPM) + Ruby Enterprise 1.8.7 + Passenger 2.2.5 + Rails 2.2 + Hello
    \n(date;free -m;ab -n 20000 -c 100 127.0.0.1:48120\/hello;date;free -m)2>&1|tee ttt;echo;egrep “(buffers\/cache)|(Requests per second)” ttt
    \nTo restart Apache: \/usr\/local\/apache-worher\/bin>.\/apachectl restart
    \n<\/span><\/p><\/blockquote>\n

      Notes:<\/p>\n
    • Same as Nginx, Apache needs to be restarted after each round of test.<\/li>\n<\/ul>\n\n\n\n\n\n\n
      Round<\/td>\nFree Memory (M)
      \nBefore<\/td>\n      		Free Memory (M)
      \nAfter<\/td>\n      		Memory Usage
      \n( Before – After)<\/td>\n      		Request\/Second
      \n[#\/sec](mean)<\/td>\n<\/tr>\n
      1<\/td>\n861<\/td>\n681<\/td>\n180M<\/td>\n201.12<\/td>\n<\/tr>\n
      2<\/td>\n860<\/td>\n682<\/td>\n153M<\/td>\n199.88<\/td>\n<\/tr>\n
      3<\/td>\n861<\/td>\n683<\/td>\n178M<\/td>\n200.73<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n","protected":false},"excerpt":{"rendered":"

      To perform a fair benchmark requiring a lot of preparation, as you need to know the detail of each product and set appropriate configuration. Recently, we got chance to do some internal testing.<\/p>\n","protected":false},"author":1,"featured_media":398,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[16,27,20],"tags":[37,196,218,243,248,255,271,294],"class_list":["post-97","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-benchmarks","category-web-server","category-performance","tag-apache-compatible","tag-nginx","tag-passenger","tag-rails","tag-ree","tag-ruby","tag-speed","tag-upload"],"jetpack_featured_media_url":"https:\/\/lswp.store\/wp-content\/uploads\/2013\/02\/Unknown.jpeg","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/posts\/97","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/comments?post=97"}],"version-history":[{"count":0,"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/posts\/97\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/media\/398"}],"wp:attachment":[{"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/media?parent=97"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/categories?post=97"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lswp.store\/index.php\/wp-json\/wp\/v2\/tags?post=97"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}