Hi,
I’m using stub_status module as a source collect metrics from my nginx
servers. I’m doing this:
(1) wget the stats page
(2) wait 1 second
(3) wget the stats page again
(4) Get the difference of the requests value from (3) and (1) to get
requests per second.
Am I doing right? I am also collecting Active Connections and want to
understand which of these values will give me a better idea on what’s
happening in my servers during high load/traffic hours.
Thanks!
Héctor Paz
Hector Danniel Paz Trillo wrote:
Am I doing right? I am also collecting Active Connections and want to
understand which of these values will give me a better idea on what’s
happening in my servers during high load/traffic hours.
For seeing nice graphs I use Munin with the nginx modules installed. You
can see attached for sample of daily and weekly graphs. There is also a
module for nginx memory usage.
For up to the second information I use the python script referenced at
http://hostingfu.com/article/quick-nginx-status-script. This will give
data like this:
Conn Conn/s Request/s Read Write Wait
486 21.80 47.90 152 75 259
476 17.70 59.80 149 54 273
448 18.20 43.90 133 56 259
428 18.30 46.90 124 68 236
459 19.00 49.62 139 63 256
Thanks!
I’m using ganglia and made my own script to get stats from nginx and
feed them to ganglia. I see that my approach is similar to the python
script so I think i’m doing right. Both the munin and python scripts
gave me some ideas to improve my script and graphs.
Regards
Hector Paz