Hello zepolen,

On Friday 28 August 2009 19:59:41 zepolen wrote:
> What's wrong with parsing the logs? That's one of the reasons they exist.

Seeing the efficiency of graphing the request/connection rate with stub
module, I think this is the way to go.
Parsing the logs can be an heavy task with high traffic.

I will check what it takes to write a module for that.


>
> On Fri, Aug 28, 2009 at 8:10 PM, Pior Bastida<pior@pbastida.net> wrote:
> > Hello All,
> >
> > There is an interesting metric about webserver: the rate of the http
> > status code.
> >
> > I would be able to plot "good"/2xx-3xx and "bad"/4xx-5xx status code
> > rate.
> >
> > Is there an efficient way to achieve that ?
> > I don't want to parse the logs of nginx.
> >
> > Thanks for any pointers
> >
> > --
> > Pior Bastida
> > pior@pbastida.net



--
Pior Bastida
pior@pbastida.net