Re: Report Generator info & Object Life Cycle

On Wed, 27 Mar 2002, Peter Backx wrote:

> I get the following warning: "Object Life Cycle birthdaysettings
> are depricated and ignored", which is possible, since my pg-files
> were last used with version 1.5 of polygraph. However in the docs
> there is no mention that this feature was removed or what the
> solution might be. I suppose I can just drop the birthday field?

See change.log. Yes, you can just delete the birthday field.
 
> Another problem is I've been unable to locate documentation on the
> Report Generator. If it exists, could some one please give me a
> link to it (you may ignore the rest of the message in that case :)
> ).

        http://www.web-polygraph.org/docs/userman/reporting.html

Some object names are explained at

        http://polygraph.ircache.net/doc/lxopts.html#--objects
 
> What I'm looking for is an explenation of some of the tables and
> graphics: * What is the "offered hit ratio" and how is it
> calculated. Is this based on some "ideal" cache? Or is it
> something else.

Offered hit is a repeated basic (simple GET) request for cachable
content. Yes, an ideal infinite cache should return a hit for each hit
offered.

> * Is the "rep.rate" the reply rate?

Yes.

> Does this mean that in the first graphic (req.rate & rep.rate vs.
> time) both curves should be the (almost) same for a correct
> cache/proxy?

Yes. If the curves differ, the proxy is probably delaying the
responses (and the delay increases).

> * What are the xact.level.mean & wait.level.mean graphics?

See lxopts.html URL above and

        http://www.web-polygraph.org/cgi-bin/htsearch?words=xact.level.mean
        http://www.web-polygraph.org/mail-archive/users/200007/0011.html
        http://polygraph.ircache.net/cgi-bin/mail2html.pl?fnames=20000601/710

Thank you for using Web Polygraph.

Alex.
Received on Wed Mar 27 2002 - 07:57:55 MST