On 6/21/07, Will Johnson <wjohnson getconnected.com>
wrote:
> >This produces log messages that look like this:
> >INFO: /select q=solr&wt=python&indent=on
hits=1 0 94
> >
> >If there was no DocSet, it would look like this:
> >INFO: /select q=solr&wt=python&indent=on -
0 94
>
> I would think that tacking the new stats onto the end
of the line would
> be better than in the middle. Usually when I parse log
files it
> involves something like:
>
> String[] arr = line.split(" ");
> code = arr[3]
> time = arr[4]
>
> instead of the following which is what it seems you're
implying that
> people are doing:
>
> String[] arr = line.splti(" ")
> code = arr[arr.length-2]
> time = arr[arr.length-1]
>
> but then again, I don't have any code written to parse
things yet so
> backwards compatibility isn't an issue for me and
either format is fine.
I do have some code, and for times it just went from the
end. The
start can vary based on the container config...the resin
logs like
this:
[2007-06-21 04:15:58.195]
rows=1&start=0&q=contentId:0 0 1
w/o backward compatibility concerns, which looks best? I
don't have a
strong opinion, but I think time at the end makes most sense
(maybe
because that's what I'm used to seeing in other logs?)
-Yonik
|