From: "Rich" 

This is a multi-part message in MIME format.

------=_NextPart_000_01D2_01C31C7B.F21D7DC0
Content-Type: text/plain;
        charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

   Again Mike Miller puts his foot in his mouth.  The Netcraft reference =
was for Intel's web site which had a pre-release version of Windows =
Server 2003 in the server pool.  As for Intel's actions, the four month =
old stale post from Netcraft, Netcraft gave several suggestions why they =
saw strange behavior.

   In the mean time, if you check the current performance of the same =
site using the link provided by netcraft, =
http://uptime.netcraft.com/up/performance?explain=3D0&mode_p=3Don&mode_u=3D=
off&mode_w=3Doff&by=3Dcollector&errors=3D0&site=3Dwww.intel.com&site1=3D&=
sample=3D8&submit=3DExamine&range=3D5d&maxy=3D0, the current
performance = from London has changed.  It's also the case that the quoted
performance = from London was the worst and more irregular than elsewhere. 
The DNS = performance from London was horrible.  Across the board the Texas
and NY = performance numbers are lower and more regular further supporting
= Netcraft's suggest that the behavior is a configuration issue.

   For those that value standard benchmarks, Windows Server 2003 is from =
64% to 355% faster than Windows 2000 Server on WebBench.  See =
http://www.veritest.com/clients/reports/microsoft/ms_performance.pdf.

Rich

  "Mike '/m'"  wrote in message =
news:mjqccvsmud5skigb379l04hn914q3nndiu{at}4ax.com...


  1) Microsoft has already shown time and time again that they do not =
share
  your discipline and caution when they set up their servers.

  2) Within each version, the consistency of the timings hint that the
  servers may be a bit more consistent.

   /m



  On Sat, 17 May 2003 12:20:49 -0400, "Geo."  wrote:

  >I don't know about everyone else but when I get a new version I =
always avoid
  >putting it on my biggest baddest production boxes till I'm sure it =
can be
  >trusted (which usually doesn't happen until after sp1 or sp2).
  >
  >Geo.
  >
  >"Mike '/m'"  wrote in message
  >news:ebbccv8q0rn8lmu7pr68sdvg519c6ckbsi{at}4ax.com...
  >>
  >> Why would they install the new OS on slower machines?
  >>
  >>  /m
  >>
  >>
  >> On Fri, 16 May 2003 19:48:41 -0400, "Geo."
 wrote:
  >>
  >> >Kinda meaningless without knowing if it's identical hardware.
  >> >
  >> >Geo.
  >> >
  >> >"Joe Barr"  wrote in message
  >> >news:pan.2003.05.15.20.43.52.824709{at}austin.rr.com...
  >> >>
  >> >> Netcraft has some interesting stats/observations about that.
  >> >>
  >> >>
  >> >>
  >> >>
  >>
  =
>>http://news.netcraft.com/archives/2003/01/21/performance_of_wwwintelcom=
_att
  >racting_interest.html

------=_NextPart_000_01D2_01C31C7B.F21D7DC0
Content-Type: text/html;
        charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable








   Again
Mike Miller puts his =
foot in his=20
mouth.  The Netcraft reference was for Intel's web site which had a =

pre-release version of Windows Server 2003 in the server pool.  As = for=20
Intel's actions, the four month old stale post from Netcraft, Netcraft = gave=20
several suggestions why they saw strange behavior.
 
   In the
mean time, if you =
check the=20
current performance of the same site using the link provided by = netcraft, http://uptime.netcraft.com/up/performance?explain=3D0&mode_p=3D=
on&mode_u=3Doff&mode_w=3Doff&by=3Dcollector&errors=3D0&am=
p;site=3Dwww.intel.com&site1=3D&sample=3D8&submit=3DExamine&a=
mp;range=3D5d&maxy=3D0">http://uptime.netcraft.com/up/performance?exp=
lain=3D0&mode_p=3Don&mode_u=3Doff&mode_w=3Doff&by=3Dcolle=
ctor&errors=3D0&site=3Dwww.intel.com&site1=3D&sample=3D8&=
amp;submit=3DExamine&range=3D5d&maxy=3D0,=20
the current performance from London has changed.  It's also the =
case that=20
the quoted performance from London was the worst and more irregular than =

elsewhere.  The DNS performance from London was
horrible.  = Across the=20
board the Texas and NY performance numbers are lower and more regular = further=20
supporting Netcraft's suggest that the behavior is a
configuration=20 issue.
 
   For those
that value =
standard=20
benchmarks, Windows Server 2003 is from 64% to 355% faster than Windows = 2000=20
Server on WebBench.  See http://www.veritest.com/clients/reports/microsoft/ms_performance.=
pdf">http://www.veritest.com/clients/reports/microsoft/ms_performance.pdf=
.
 
Rich
 

  "Mike '/m'" <mike{at}barkto.com>=20">mailto:mike{at}barkto.com">mike{at}barkto.com>=20
  wrote in message news:mjqccvsmud5=
skigb379l04hn914q3nndiu{at}4ax.com...1)=20
  Microsoft has already shown time and time again that they do not =
shareyour=20
  discipline and caution when they set up their servers.2) =
Within each=20
  version, the consistency of the timings hint that theservers may =
be a bit=20
  more
consistent. /mOn
Sat, 17 May 2003 =
12:20:49=20
  -0400, "Geo." <georger{at}nls.net>=20">mailto:georger{at}nls.net">georger{at}nls.net>=20
  wrote:>I don't know about everyone else but
when I get a =
new=20
  version I always avoid>putting it on my biggest baddest =
production=20
  boxes till I'm sure it can be>trusted (which usually doesn't =
happen=20
  until after sp1 or
sp2).>>Geo.>>"Mike
=
'/m'" <mike{at}barkto.com>">mailto:mike{at}barkto.com">mike{at}barkto.com>
wrote in=20
  =
message>news:ebbccv8q0rn8lmu7pr68sdvg519c6ckbsi{at}4ax.com...>=
>>>=20
  Why would they install the new OS on slower=20
  machines?>>>>  =
/m>>>>>>=20
  On Fri, 16 May 2003 19:48:41 -0400, "Geo." <georger{at}nls.net>=20">mailto:georger{at}nls.net">georger{at}nls.net>=20
  wrote:>>>> >Kinda
meaningless without knowing =
if it's=20
  identical hardware.>>
>>>
>Geo.>>=20
  >>> >"Joe Barr" <warthawg{at}austin.rr.com>">mailto:warthawg{at}austin.rr.com">warthawg{at}austin.rr.com>
=
wrote in=20
  message>>=20
 
>news:pan.2003.05.15.20.43.52.824709{at}austin.rr.com...>>=20
  >>>> >> Netcraft has
some interesting =
stats/observations=20
  about that.>>
>>>>
>>>>=20
  >>>>=20
  =
>>>>>>http://news.netcraft.com/archives/2003/01=
/21/performance_of_wwwintelcom_att>racting_interest.html

------=_NextPart_000_01D2_01C31C7B.F21D7DC0--

--- BBBS/NT v4.01 Flag-4