Re: 2.5.STABLE13 approaching

From: Guido Serassio <guido.serassio@dont-contact.us>
Date: Mon, 27 Feb 2006 17:26:17 +0100

Hi Henrik,

At 16.55 27/02/2006, Henrik Nordstrom wrote:
> > - For standard helpers, the displayed value is the length of the last
> > request, and not now - time_when_last_request_was_sent
>
>Are you sure? Hmm.. code confirms what you say. Changed by Duane
>15-May-02..
>
>revision 1.34.2.3
>date: 2002/05/15 19:09:53; author: wessels; state: Exp; lines: +7 -5
>MFC: Steve Bennett fixes for helper/redirector service time values.
>
>- storeAppendPrintf(sentry, "avg service time: %d msec\n",
>- hlp->stats.avg_svc_time);
>+ storeAppendPrintf(sentry, "avg service time: %.2f msec\n",
>+ (double) hlp->stats.avg_svc_time / 1000.0);
>
>- tt = 0.001 * tvSubMsec(srv->dispatch_time, current_time);
>+ tt = 0.001 * tvSubMsec(srv->dispatch_time,
>+ srv->flags.busy ? current_time : srv->answer_time);
>
>
>now to display just what you say, probably with good intentions.
>
>So why do we want to change this back? Isn't it good to know how long
>the response took? Never really saw that "idle" time as useful.

Sure, is good, but is not the same meaning as stateful helpers, and
this is very confusing.
I think that both info are useful. But, probably, to avoid any new
problem, the best thing for 2.5 is to leave the current behaviour,
fixing only the unused helper problem.

But in 3.0 we should rearrange the helper stats, now they are too confusing.

Regards

Guido

-
========================================================
Guido Serassio
Acme Consulting S.r.l. - Microsoft Certified Partner
Via Lucia Savarino, 1 10098 - Rivoli (TO) - ITALY
Tel. : +39.011.9530135 Fax. : +39.011.9781115
Email: guido.serassio@acmeconsulting.it
WWW: http://www.acmeconsulting.it/
Received on Mon Feb 27 2006 - 09:26:26 MST

This archive was generated by hypermail pre-2.1.9 : Tue Feb 28 2006 - 12:00:05 MST