Boot Delay

In news:jncftt$ltb$,
Kernelbugger wrote:

I'm unable to figure out why one of my HDDs takes only two sweeps of
the bullets across the WindowsXP screen during boot up, and another
takes up to 15 sweeps. I'm referring to the blue buttons that sweep
horizontally right after POST. I ran fixmbr and fixboot and the
number of sweeps was unchanged. I plan on running boot log to look
for the culprit, but I thought I had it figured because the two-sweep
HDD was on SATA port 2 and the 15 sweeper was on SATA port 5. I
swapped which port the drives were on, and the 15-sweeper now takes
only 2 sweeps. Next I cloned the 2-sweeper and the clone took 15
sweeps to boot on SATA port 2. I'd like a reference to what Windows
XP is doing when it first starts to load after the POST. Or is that
exactly what bootlog will tell me? TIA
Asrock 890FX Deluxe 4, 4-core 3.6 GB CPU, 2x2 MB 1600 MHz RAM,
Windows XP Pro, and HDDs of 500 and 1 TB.

Whoa wait a minute! Sweeps of the bullets across the Windows XP screen?
I think I know what you are talking about and I don't think that counts
for anything. As lots of things can change that and that doesn't mean a
problem. As I think the only purpose is to show it isn't hung up and the
computer is frozen. So any movement means it is still working and
everything is good (or good and slow).

What does help is how long it takes to boot up. Although this is often
misleading as well. Some count when the desktop pops up. Others like
using something else. I like using the time when the disk activity stops
to almost nil and the CPU usage drops to idle. This time is the longest
of the longest of course.

So how long does it take before everything is snappy like you had it
running at idle for 10-15 minutes or so? Then we are getting somewhere.

Bill
Gateway M465e ('06 era) - OE-QuoteFix v1.19.2
Centrino Core Duo T2400 1.83GHz - 2GB - Windows XP SP3

From power on to the desktop is slightly under one minute. I'm not having
a problem, it's just that I'm a perfectionist. I built the machine.

I can't figure out if one SATA 3 port is faster than another, as it appears,
or what may be the cause. I ran bootlog, and it shows system 13.9 secs,
explorer 6.3 secs, and nusb3mon.exe 7.3 secs. All the other items were
loaded simultaneously it appears, and each took only a few seconds.




"BillW50" wrote in message
news:jnci93$2i3$

In news:jncftt$ltb$,
Kernelbugger wrote:

I'm unable to figure out why one of my HDDs takes only two sweeps of
the bullets across the WindowsXP screen during boot up, and another
takes up to 15 sweeps. I'm referring to the blue buttons that sweep
horizontally right after POST. I ran fixmbr and fixboot and the
number of sweeps was unchanged. I plan on running boot log to look
for the culprit, but I thought I had it figured because the two-sweep
HDD was on SATA port 2 and the 15 sweeper was on SATA port 5. I
swapped which port the drives were on, and the 15-sweeper now takes
only 2 sweeps. Next I cloned the 2-sweeper and the clone took 15
sweeps to boot on SATA port 2. I'd like a reference to what Windows
XP is doing when it first starts to load after the POST. Or is that
exactly what bootlog will tell me? TIA
Asrock 890FX Deluxe 4, 4-core 3.6 GB CPU, 2x2 MB 1600 MHz RAM,
Windows XP Pro, and HDDs of 500 and 1 TB.

Whoa wait a minute! Sweeps of the bullets across the Windows XP screen? I
think I know what you are talking about and I don't think that counts for
anything. As lots of things can change that and that doesn't mean a
problem. As I think the only purpose is to show it isn't hung up and the
computer is frozen. So any movement means it is still working and
everything is good (or good and slow).

What does help is how long it takes to boot up. Although this is often
misleading as well. Some count when the desktop pops up. Others like using
something else. I like using the time when the disk activity stops to
almost nil and the CPU usage drops to idle. This time is the longest of
the longest of course.

So how long does it take before everything is snappy like you had it
running at idle for 10-15 minutes or so? Then we are getting somewhere.

Bill
Gateway M465e ('06 era) - OE-QuoteFix v1.19.2
Centrino Core Duo T2400 1.83GHz - 2GB - Windows XP SP3

In news:jncoj8$6g6$,
Kernelbugger wrote:

From power on to the desktop is slightly under one minute. I'm not
having a problem, it's just that I'm a perfectionist. I built the
machine.
I can't figure out if one SATA 3 port is faster than another, as it
appears, or what may be the cause. I ran bootlog, and it shows
system 13.9 secs, explorer 6.3 secs, and nusb3mon.exe 7.3 secs. All
the other items were loaded simultaneously it appears, and each took
only a few seconds.

Oh man! As good as I think I am I might be able to tweak it myself a
hair better (at best a second or two), but why? Even under the fastest
SSD commercially available you still won't break 30 seconds (at least I
don't think I can break that limit).

"BillW50" wrote in message
news:jnci93$2i3$

In news:jncftt$ltb$,
Kernelbugger wrote:

I'm unable to figure out why one of my HDDs takes only two sweeps of
the bullets across the WindowsXP screen during boot up, and another
takes up to 15 sweeps. I'm referring to the blue buttons that sweep
horizontally right after POST. I ran fixmbr and fixboot and the
number of sweeps was unchanged. I plan on running boot log to look
for the culprit, but I thought I had it figured because the
two-sweep HDD was on SATA port 2 and the 15 sweeper was on SATA
port 5. I swapped which port the drives were on, and the
15-sweeper now takes only 2 sweeps. Next I cloned the 2-sweeper
and the clone took 15 sweeps to boot on SATA port 2. I'd like a
reference to what Windows XP is doing when it first starts to load
after the POST. Or is that exactly what bootlog will tell me? TIA
Asrock 890FX Deluxe 4, 4-core 3.6 GB CPU, 2x2 MB 1600 MHz RAM,
Windows XP Pro, and HDDs of 500 and 1 TB.

Whoa wait a minute! Sweeps of the bullets across the Windows XP
screen? I think I know what you are talking about and I don't think
that counts for anything. As lots of things can change that and that
doesn't mean a problem. As I think the only purpose is to show it
isn't hung up and the computer is frozen. So any movement means it
is still working and everything is good (or good and slow).

What does help is how long it takes to boot up. Although this is
often misleading as well. Some count when the desktop pops up.
Others like using something else. I like using the time when the
disk activity stops to almost nil and the CPU usage drops to idle.
This time is the longest of the longest of course.

So how long does it take before everything is snappy like you had it
running at idle for 10-15 minutes or so? Then we are getting
somewhere. --
Bill
Gateway M465e ('06 era) - OE-QuoteFix v1.19.2
Centrino Core Duo T2400 1.83GHz - 2GB - Windows XP SP3

Bill
Gateway M465e ('06 era) - OE-QuoteFix v1.19.2
Centrino Core Duo T2400 1.83GHz - 2GB - Windows XP SP3

Kernelbugger wrote:

I'm unable to figure out why one of my HDDs takes only two sweeps of the
bullets across the WindowsXP screen during boot up, and another takes up to
15 sweeps. I'm referring to the blue buttons that sweep horizontally right
after POST. I ran fixmbr and fixboot and the number of sweeps was
unchanged. I plan on running boot log to look for the culprit, but I
thought I had it figured because the two-sweep HDD was on SATA port 2 and
the 15 sweeper was on SATA port 5. I swapped which port the drives were on,
and the 15-sweeper now takes only 2 sweeps. Next I cloned the 2-sweeper and
the clone took 15 sweeps to boot on SATA port 2. I'd like a reference to
what Windows XP is doing when it first starts to load after the POST. Or
is that exactly what bootlog will tell me? TIA

Asrock 890FX Deluxe 4, 4-core 3.6 GB CPU, 2x2 MB 1600 MHz RAM, Windows XP
Pro, and HDDs of 500 and 1 TB.

SB850 6 ports SATA3
Marvell SE9123/9120 2 ports SATA3
1 ata133 IDE connector (on Marvell maybe???)

SB850 Databook (PDF page 36)
http://support.amd.com/us/Embedded_.../44758.pdf

"The SATA controller can operate in three modes:

* All six channels are configured as SATA AHCI mode.

* All six channels are configured as IDE mode. In this configuration,
the SATA controller is configured into two IDE controllers, with the
programming interface of channel 0 to 3 under the first IDE controller,
and that of channel 4 and 5 under the second IDE controller.

* Four channels (channel 0 to 3) are configured as SATA AHCI and
two channels (channel 4 and 5) are configured as IDE mode. In this
configuration, the programming interface of channel 4 and 5 are
under the IDE controller."

*******

It's possible some AHCI code is slower than IDE code, under the
same conditions. Switching ports, might be causing a different
driver to be used (either by the BIOS Extended INT 0x13 read
code, or by a driver in the OS). I don't know how a theory like
this pans out though, with respect to what drivers you could have
installed in WinXP at the same time. Would it tolerate a mix ?
I suppose a browse of Device Manager, might tell you.

Paul

On Thu, 26 Apr 2012 21:07:57 -0500, "BillW50" wrote
in article <jncv1v$344$...

In news:jncoj8$6g6$,
Kernelbugger wrote:
> From power on to the desktop is slightly under one minute. I'm not
> having a problem, it's just that I'm a perfectionist. I built the
> machine.
> I can't figure out if one SATA 3 port is faster than another, as it
> appears, or what may be the cause. I ran bootlog, and it shows
> system 13.9 secs, explorer 6.3 secs, and nusb3mon.exe 7.3 secs. All
> the other items were loaded simultaneously it appears, and each took
> only a few seconds.

Oh man! As good as I think I am I might be able to tweak it myself a
hair better (at best a second or two), but why? Even under the fastest
SSD commercially available you still won't break 30 seconds (at least I
don't think I can break that limit).

I don't know about XP, but Windows 7 on a fast SSD boots to the desktop
in under 30 seconds easily, without any performance tweaking. My
production system at work achieves this. Normally I have to log on,
which of course adds to the time, but I tested with auto logon enabled
and got 25 seconds pretty consistently. It's not a particularly
special PC - Dell Optiplex 790, Core i5-2500 at 3.3GHz with 8GB RAM
running Windows 7 Enterprise 64-Bit on a Corsair Force 3 SSD. Our
developers have much more powerful machines, I'm tempted to see what
theirs can do...

Zaphod

The secret of flying is to hurl yourself to the ground, and miss.