Hi,

can anybody tell me how a long the Sun Blade 2500 and/or 1500
takes to initialize? I mean the time from pressing the power botton
in power-off state until the system starts booting. The total
time until a SX10b72 release is up and running would also be
interesting.

Is it as awful long as the Blade 1000 takes or is it at least
as fast as a PeeCee?


Cheers,

Tom

Thomas Maier-Komor wrote:
> Hi,
>
> can anybody tell me how a long the Sun Blade 2500 and/or 1500
> takes to initialize? I mean the time from pressing the power botton
> in power-off state until the system starts booting. The total
> time until a SX10b72 release is up and running would also be
> interesting.
>
> Is it as awful long as the Blade 1000 takes or is it at least
> as fast as a PeeCee?

Do you have SCSI on your PC?? If so, then you know it will
take a long time to boot up.

Usually Sun high end workstations are not used as embedded
or set top appliances.... why the need for ultra fast starts?

It can what seem forever for my Dual 3.2 Xeon to boot due to
it's hardware raid controller. My dual Opteron takes a long time
too... and it just had dual channel scsi. In all fairness, I think
the SunBlade is faster than both.

Chris Cox wrote:
> Thomas Maier-Komor wrote:
>
>> Hi,
>>
>> can anybody tell me how a long the Sun Blade 2500 and/or 1500
>> takes to initialize? I mean the time from pressing the power botton
>> in power-off state until the system starts booting. The total
>> time until a SX10b72 release is up and running would also be
>> interesting.
>>
>> Is it as awful long as the Blade 1000 takes or is it at least
>> as fast as a PeeCee?
>
>
> Do you have SCSI on your PC?? If so, then you know it will
> take a long time to boot up.
>
> Usually Sun high end workstations are not used as embedded
> or set top appliances.... why the need for ultra fast starts?
>

Well, I just thought of the very long time a Blade 1000 takes
to do its power-on self tests. After pressing the power button of
this machine it takes about a minute for the frame-buffer to
be activated and some more time to pass all the remaing tests.

Of cource a SCSI controller takes a while to get ready. I also have
some (non-RAID), which, if configured to probe only for existing
devices, are ready within few seconds.

You are right, ultra fast starts are not required. But having to wait
about two to three minutes for the machine to become ready is an
obvious drawback, if its main use is at home where you normally
turn it off after use to save some energy.

> It can what seem forever for my Dual 3.2 Xeon to boot due to
> it's hardware raid controller. My dual Opteron takes a long time
> too... and it just had dual channel scsi. In all fairness, I think
> the SunBlade is faster than both.
>

IIRC, the B2500 does only have a single SCSI controller and one
IDE controller. So your comparison implies that the power on self
tests must have become faster compared to the B1000 and are now at
least on par with PCs. Right?

Thanks,

Tom

Thomas Maier-Komor <maierkom@lpr.e-technik.no-spam.tu-muenchen.de> wrote:
> Well, I just thought of the very long time a Blade 1000 takes
> to do its power-on self tests. After pressing the power button of
> this machine it takes about a minute for the frame-buffer to
> be activated and some more time to pass all the remaing tests.
>
> You are right, ultra fast starts are not required. But having to wait
> about two to three minutes for the machine to become ready is an
> obvious drawback, if its main use is at home where you normally
> turn it off after use to save some energy.

The time taken to the the POST is obviously an annoyance, but so is a
system which doesn't detect a faulty component during it's POST, and thus
crahes further down the track (or even worse, corrupts data).

You can configure the level of testing done, although what you're seeing
is probably the minimum (a few minutes is fairly fast - it's possible to
configure some systems to take many hours to POST, although you'd only
normally do that when they are new/relocated/after a major problems/etc)

Scott

Ever waited for an RS6000 before??

Now... that's a wait!


I can't say if the SunBlade 2500 is as fast as a PC booting because there are
too many variables. My guess is that it's easier to tweak the PC to be
faster at bootup in all cases... but you get a lot less.

Chris Cox schrieb:
> Ever waited for an RS6000 before??
>
> Now... that's a wait!
>
>
> I can't say if the SunBlade 2500 is as fast as a PC booting because
> there are
> too many variables. My guess is that it's easier to tweak the PC to be
> faster at bootup in all cases... but you get a lot less.

Sounds reasonable - like always it seems to be:
"you get, what you pay for".

This time we just "pay" some seconds being idle waiting for the
machine... At least it seems to be worthwhile as nobody owning
such a machine cried out that it is a PITA. Or does nobody have
such a beast?!


Thanks,

Tom

Thomas Maier-Komor <maierkom@lpr.e-technik.no-spam.tu-muenchen.de> writes:

> Chris Cox schrieb:
> > Ever waited for an RS6000 before??
> > Now... that's a wait!
> >
> > I can't say if the SunBlade 2500 is as fast as a PC booting because
> > there are
> > too many variables. My guess is that it's easier to tweak the PC to be
> > faster at bootup in all cases... but you get a lot less.
>
> Sounds reasonable - like always it seems to be:
> "you get, what you pay for".
>
> This time we just "pay" some seconds being idle waiting for the
> machine... At least it seems to be worthwhile as nobody owning
> such a machine cried out that it is a PITA. Or does nobody have
> such a beast?!

I have an SB1000 and an SB2500 together here at work. The older
machine seems to take longer to start. They are both faster than my PC
due to the amount of invasive virus protection, inventory management
and general fluff they load on our PCs, so actually I don't see a
problem (I love my unix machines, they keep me sane).

Chris
--
Chris Morgan
"Post posting of policy changes by the boss will result in
real rule revisions that are irreversible"

- anonymous correspondent

Chris Morgan wrote:
>
> I have an SB1000 and an SB2500 together here at work. The older
> machine seems to take longer to start. They are both faster than my PC
> due to the amount of invasive virus protection, inventory management
> and general fluff they load on our PCs, so actually I don't see a
> problem (I love my unix machines, they keep me sane).
>
> Chris

Sounds good - so only few drawbacks remain. But I guess there are always
some: Insanity on windows side, 90% products on linux side, and
unsupported hardware on solaris side.

Now I hope some weird reason comes to my mind to justify the bags of
money such a beast costs to buy ;-)

Tom

On Thu, 6 Jan 2005, Thomas Maier-Komor wrote:

> Now I hope some weird reason comes to my mind to justify the bags of
> money such a beast costs to buy ;-)

How about "because"? :-)

--
Rich Teer, SCNA, SCSA, author of "Solaris Systems Programming"

. * * . * .* .
. * . .*
President, * . . /\ ( . . *
Rite Online Inc. . . / .\ . * .
.*. / * \ . .
. /* o \ .
Voice: +1 (250) 979-1638 * '''||''' .
URL: http://www.rite-online.net ******************

In article <cragdc$s4p@library2.airnews.net>, Chris Cox wrote:
>Ever waited for an RS6000 before??

Having had to babysit a dodgy RS/6000 S-70 rebooting, I can assure you
that I never, ever want to deal with that fscker again. Sit around
waiting for the thing to come up, watching the random numbers
incrementing on the little front panel.

--
Andre.

Email: andre at africa dot purplecow dot org.