Graph nginx by error codes and requests per second? Cacti? or some other?

Server NGINX
1

Does anybody have a monitoring system in place by nginx error code…
500, 200, 404, 444… and did you do this with cacti or php4nagios?

Regards,
Joseph

2

On Tue, Jun 5, 2012 at 10:26 AM, Joseph C. [email protected]
wrote:

Does anybody have a monitoring system in place by nginx error code… 500,
200, 404, 444… and did you do this with cacti or php4nagios?

Hi,

You can take a look at the nginx-lua module (on the logby branch) :
https://github.com/chaoslawful/lua-nginx-module/tree/logby

There is an example in the README :
https://github.com/chaoslawful/lua-nginx-module/blob/logby/README
Look for log_by_lua, and log_by_lua_file.

You can use it to aggregate values, and use another location to report
aggregated data (using content_by_lua) and feed it in your own system.

We use OpenTSDB (http://opentsdb.net/) to keep aggregating data in time
series.

Hope that helps,

Matthieu.

3

matthieu,

Thanks! If you are using it to graph nginx by error code… would you
have any screenshot example without any sensitive data that you can
share of the final outcome of this? Iam no way familiar with the things
you quoted.

Anybody know an alternative?

Regards,

Joseph

4

Hello!

On Wed, Jun 6, 2012 at 2:26 AM, Matthieu T. >

You can take a look at the nginx-lua module (on the logby branch) :
https://github.com/chaoslawful/lua-nginx-module/tree/logby

BTW, The “logby” branch is going to be merged into “master” in the
next few days or so (after more aggressive testing) :slight_smile:

Special thanks go to Matthieu T. for the work on this branch :slight_smile:

Best regards,
-agentzh