CentOS Version: CentOS 7 Centmin Mod Version Installed: Centmin Mod 1.2.3-eva2000.08 Nginx Version Installed: 1.11.10 PHP Version Installed: 7.0.0 - Zend opcache v7.0.6-dev Persistent Config: No. As per our chat on lowendtalk, i am having some issues when under high load. I have installed wordpress via option 22 and went in and disabled all plugins; changed theme to default wordpress theme; enabled w3 total cache. This is the result of centmin under load: Screenshot This is the EXACT same VM from Vultr with same spec, standard wordpress install with same plugins and theme. The result is different: Screenshot
Best place to ask as pretty sure it ain't normal for Centmin Mod installed Wordpress to average 8+ seconds per page load What are your server specs for cpu cores, ram, disk and web host ? Location of your server geographically ? loader.io test parameters used ? as wordpress was installed via centmin.sh menu option 22, you would of got outputted the path location where it will create the domain name's vhost conf file named newdomain.com.conf (and newdomain.com.ssl.conf if you selected yes to self signed SSL) Nginx vhost conf path will be at /usr/local/nginx/conf/conf.d/newdomain.com.conf Nginx HTTP/2 SSL vhost conf path will be at /usr/local/nginx/conf/conf.d/newdomain.com.ssl.conf Nginx Self-Signed SSL Certificate Directory at /usr/local/nginx/conf/ssl/newdomain.com Vhost public web root will be at /home/nginx/domains/newdomain.com/public Vhost log directory will be at /home/nginx/domains/newdomain.com/log Please post the contents of /usr/local/nginx/conf/conf.d/newdomain.com.conf and if applicable /usr/local/nginx/conf/conf.d/newdomain.com.ssl.conf wrapped in CODE tags (outlined at How to use forum BBCODE code tags) which wp plugins did you remove ? and which wp plugins did you add yourself ? You can find out via the below commands too which wp plugins are installed where newdomain.com is yourdomain.com name if you installed wordpress in web root Code (Text): cd /home/nginx/domains/newdomain.com/public wp plugin status --allow-root and post output wrapped in CODE tags.
oh i see you're using 123.08stable Centmin Mod, it's centmin.sh menu option 22 isn't as optimised as 123.09beta01 branch too. But still 8+ seconds isn't normal. if you want to double check, you can do a fresh CentOS 7 install and try 123.09beta01 install Centmin Mod LEMP Stack Install Nginx on CentOS Code (Text): yum -y update; curl -O https://centminmod.com/betainstaller.sh && chmod 0700 betainstaller.sh && bash betainstaller.sh Or with PHP 7 default as per https://community.centminmod.com/threads/centmin-mod-09-beta-branch-testing.4128/ Code (Text): yum -y update; curl -O https://centminmod.com/betainstaller7.sh && chmod 0700 betainstaller7.sh && bash betainstaller7.sh then run centmin.sh menu option 22 for wordpress install
Code: [root@syd2 public]# cat /usr/local/nginx/conf/conf.d/vultr.domain.net.au.conf # Centmin Mod Getting Started Guide # must read http://centminmod.com/getstarted.html # redirect from non-www to www # uncomment, save file and restart Nginx to enable # if unsure use return 302 before using return 301 #server { # listen 80; # server_name vultr.domain.net.au; # return 301 $scheme://www.vultr.domain.net.au$request_uri; # } server { server_name vultr.domain.net.au www.vultr.domain.net.au; # ngx_pagespeed & ngx_pagespeed handler #include /usr/local/nginx/conf/pagespeed.conf; #include /usr/local/nginx/conf/pagespeedhandler.conf; #include /usr/local/nginx/conf/pagespeedstatslog.conf; # limit_conn limit_per_ip 16; # ssi on; access_log /home/nginx/domains/vultr.domain.net.au/log/access.log combined buffer=256k flush=60m; error_log /home/nginx/domains/vultr.doamin.net.au/log/error.log; root /home/nginx/domains/vultr.dmaoin.net.au/public; include /usr/local/nginx/conf/wpsupercache_vultr.domain.net.au.conf; location / { # Enables directory listings when index file not found #autoindex on; # for wordpress super cache plugin try_files /wp-content/cache/supercache/$http_host/$cache_uri/index.html $uri $uri/ /index.php?q=$uri&$args; # Wordpress Permalinks #try_files $uri $uri/ /index.php?q=$uri&$args; } location ~* /(wp-login\.php) { limit_req zone=xwplogin burst=1 nodelay; #limit_conn xwpconlimit 30; auth_basic "Private"; auth_basic_user_file /home/nginx/domains/vultr.domain.net.au/htpasswd_wplogin; include /usr/local/nginx/conf/php-wpsc.conf; } location ~* /(xmlrpc\.php) { limit_req zone=xwprpc burst=45 nodelay; #limit_conn xwpconlimit 30; include /usr/local/nginx/conf/php-wpsc.conf; } d include /usr/local/nginx/conf/wpsecure_vultr.domain.net.au.conf; include /usr/local/nginx/conf/php-wpsc.conf; include /usr/local/nginx/conf/staticfiles.conf; include /usr/local/nginx/conf/drop.conf; #include /usr/local/nginx/conf/errorpage.conf; include /usr/local/nginx/conf/vts_server.conf; }
Code: 25 installed plugins: I wp-security-scan 4.0.5 I akismet 3.3 I autoptimize 2.1.0 I db-cache-reloaded-fix 2.3 I disable-xml-rpc 1.0.1 I go-newrelic 0.3 I google-analytics-for-wordpress 6.1.7 I google-authenticator 0.48 I gtmetrix-for-wordpress 0.4.2 I hello 1.6 I limit-login-attempts 1.7.1 I no-longer-in-directory 1.0.60 I p3-profiler 1.5.3.9 I query-monitor 2.13.4 I rocket-lazy-load 1.1.1 I search-regex 1.4.16 I addthis-smart-layers 3.0.0 I sucuri-scanner 1.8.3 I tpc-memory-usage 0.9.1 A w3-total-cache 0.9.5.2 I wp-optimize 2.1.1 I wp-super-cache 1.4.9 I wp-super-cache-clear-cache-menu 1.4 I wp-updates-notifier 1.4.4 I wordpress-seo 4.6 Legend: I = Inactive, A = Active
And just to be clear on something, this is under any of the high end load testing.. if I simply do individual load testing from my local connection, page loads are around 0.5-0.8s each page.
I don't have them infront of me right now as I am not infront of a computer for a few hours. However, off the top of my head it was the first option where it was spread out evenly over 1 minute and min 500 connections is where it started to die. You can do up to 10k connections per minute. Both serverpilot and centmin had exact test setup.
i assume same location too in Sydney for both serverpilot and centmin mod ? i suspect could partially be related to 123.08stable vs 123.09beta01 differences too as out of box nginx config is more optimised in 123.09beta01 as 123.08stable out of box is configured for 512 connections/s per cpu thread while 123.09beta01 is configured for now 3,000 connections/s per cpu thread
Okay interesting. I will test out once I get back. Because this only has 1 core - most likely the issue!
Yeah, i just backported 123.09beta01 nginx.conf settings to 123.08stable too so both branches should in theory support 3,000 connections/sec per cpu thread. Centmin Mod does auto optimise out of box, but end users will have to further optimise beyond that for their specific web app requirements usually. Centmin mod 123.09beta01 just has a better starting config for this + better centmin.sh menu option 22 wordpress installer
Awesome. Thanks mate. Can't wait to test it out. Curious, from the stable to beta, what are the biggest bugs which still need ironing out before going into a production environment?
honestly nothing right now - 123.09beta01 has been in development for 18+ months and only thing it's waiting for stable release is me updating all centminmod.com official documentation and faq and guides with 123.09beta01 relevant documentation as right now only the forums have most relevant info for 123.09beta01. As you can imagine that's alot of updating to do when I only work on Centmin Mod in my free time That doesn't mean there isn't any bugs, just nothing major - I've been running 120+ servers with 123.09beta01 for past 18+ months now - including this very forum and centminmod.com official site's servers even added a mention of this fact on official install page Centmin Mod LEMP Stack Install Nginx on CentOS
@joshuaharradence test results are in so pretty much sure it's related to 123.08stable vs 123.09beta01 default nginx config differences https://community.centminmod.com/th...1-load-testing-on-vultr-sydney-kvm-vps.11330/ and tip, dump w3tc plugin and use one of 123.09beta01 offered cache options, keycdn cache enabler or wp super cache or redis nginx level caching
Yep, the BETA fixed it That's awesome! And you are 100% right regarding the cache enabler plugin, made a big difference. P.S - from your own testing, which do you prefer? Cache enabler or redis caching?