What causes for the delay in connecting to an Nginx server and for the Waiting time are under an admin's control?
I've set up GTMetrix to run daily web page load speed tests, and sometimes I get surprisingly high values for fetching static files. Here's one report:
The GTMetrix analysis occurred 20 minutes ago. Our 1GB Linode web server is far from loaded. Load average: 0.17 0.06 0.06
right now, CPU utilization under 1% around the 5-minute interval during the analysis, nginx connections under 1 per second.
The nginx config file is nothing special, and hasn't changed in weeks:
server {
listen 80;
listen [::]:80;
listen 443 ssl;
listen [::]:443 ssl;
server_name analytics.idorecall.com;
ssl_certificate ...;
ssl_certificate_key ...;
root /var/www/piwik;
location / {
index index.php;
}
location ~* \.(ico|css|js|gif|png|jpe?g)$ {
add_header Access-Control-Allow-Origin idorecall.com;
access_log off;
expires max;
break;
}
location ~ \.php$ {
include snippets/fastcgi-php.conf;
fastcgi_pass unix:/var/run/php5-fpm.sock;
}
}
On the other hand, Google's Tag Manager .js also took 2 seconds to download, so maybe the problem is on GTMetrix's side? Are these random slow load times worth investigating, or should I chalk them off as a connectivity vagaries?