[gentoo-amd64] mce log errors

[gentoo-amd64] mce log errors

[Deedra Waters]

All,

I'm getting a lot of these, but it only seems to happen when i put the
machine under a lot of stress, and even then it's not always happening.
This machine is a duel opteron 242, the board is an asus k8, and with
the latest bios update, the machine has  no real problems at all.

MCE 1
CPU 0 4 northbridge TSC 8f1a7b270b6f
ADDR 75c3320
  Northbridge ECC error
  ECC syndrome = 62
       bit32 = err cpu0
       bit46 = corrected ecc error
  bus error 'local node origin, request didn't time out
      generic read mem transaction
      memory access, level generic'
STATUS 9431400100000813 MCGSTATUS 0
MCE 2
CPU 0 2 bus unit TSC 8f8ad2325db7
  L2 cache ECC error
  Bus or cache array error
       bit46 = corrected ecc error
       bit62 = error overflow (multiple errors)
  bus error 'local node origin, request didn't time out
      prefetch mem transaction
      memory access, level generic'
STATUS d000400000000863 MCGSTATUS 0


Thanks!


-- 
Deedra Waters - Gentoo developer relations, accessibility and infrastructure -
dmwaters@gentoo.org
Gentoo linux: http://www.gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] mce log errors

[Daniel Gryniewicz]

[-- Attachment #1: Type: text/plain, Size: 1151 bytes --]

On Tue, 2005-12-06 at 14:56 -0600, Deedra Waters wrote:
> All,
> 
> I'm getting a lot of these, but it only seems to happen when i put the
> machine under a lot of stress, and even then it's not always happening.
> This machine is a duel opteron 242, the board is an asus k8, and with
> the latest bios update, the machine has  no real problems at all.
> 
> MCE 1
> CPU 0 4 northbridge TSC 8f1a7b270b6f
> ADDR 75c3320
>   Northbridge ECC error
>   ECC syndrome = 62
>        bit32 = err cpu0
>        bit46 = corrected ecc error
>   bus error 'local node origin, request didn't time out
>       generic read mem transaction
>       memory access, level generic'
> STATUS 9431400100000813 MCGSTATUS 0
> MCE 2
> CPU 0 2 bus unit TSC 8f8ad2325db7
>   L2 cache ECC error
>   Bus or cache array error
>        bit46 = corrected ecc error
>        bit62 = error overflow (multiple errors)
>   bus error 'local node origin, request didn't time out
>       prefetch mem transaction
>       memory access, level generic'
> STATUS d000400000000863 MCGSTATUS 0

CPU cache is getting ECC errors.  Smells like overheating.

Daniel

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-amd64] mce log errors

[Deedra Waters]

Is there a way to test that fact? I've tried to work with lm_sensors,
but the readings for that are way way off. So, considering  lm_sensors
isuseless is there another way to tell if overheating is the problem?

The case itself has a lot of fans, but it's also got 5 harddrives in it.
On Tue, 6 Dec 2005, Daniel Gryniewicz wrote:

> Date: Tue, 06 Dec 2005 18:39:48 -0500
> From: Daniel Gryniewicz <dang@gentoo.org>
> Reply-To: gentoo-amd64@lists.gentoo.org
> To: gentoo-amd64@lists.gentoo.org
> Subject: Re: [gentoo-amd64] mce log errors
>
> On Tue, 2005-12-06 at 14:56 -0600, Deedra Waters wrote:
> > All,
> >
> > I'm getting a lot of these, but it only seems to happen when i put the
> > machine under a lot of stress, and even then it's not always happening.
> > This machine is a duel opteron 242, the board is an asus k8, and with
> > the latest bios update, the machine has  no real problems at all.
> >
> > MCE 1
> > CPU 0 4 northbridge TSC 8f1a7b270b6f
> > ADDR 75c3320
> >   Northbridge ECC error
> >   ECC syndrome = 62
> >        bit32 = err cpu0
> >        bit46 = corrected ecc error
> >   bus error 'local node origin, request didn't time out
> >       generic read mem transaction
> >       memory access, level generic'
> > STATUS 9431400100000813 MCGSTATUS 0
> > MCE 2
> > CPU 0 2 bus unit TSC 8f8ad2325db7
> >   L2 cache ECC error
> >   Bus or cache array error
> >        bit46 = corrected ecc error
> >        bit62 = error overflow (multiple errors)
> >   bus error 'local node origin, request didn't time out
> >       prefetch mem transaction
> >       memory access, level generic'
> > STATUS d000400000000863 MCGSTATUS 0
>
> CPU cache is getting ECC errors.  Smells like overheating.
>
> Daniel
>

-- 
Deedra Waters - Gentoo developer relations, accessibility and infrastructure -
dmwaters@gentoo.org
Gentoo linux: http://www.gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] mce log errors

[Deedra Waters]

Hrm, it doesn't look to be a heat issue. I opened the case and took a
look inside when i saw the last message,and it looks perfectly cool and
happy.

I do notice however, that it's only happening when i hammer the raid
array, which is on a pci promise controler.

On Tue, 6 Dec 2005, Deedra Waters wrote:

> Date: Tue, 6 Dec 2005 22:04:50 -0600 (CST)
> From: Deedra Waters <dmwaters@gentoo.org>
> Reply-To: gentoo-amd64@lists.gentoo.org
> To: gentoo-amd64@lists.gentoo.org
> Subject: Re: [gentoo-amd64] mce log errors
>
> Is there a way to test that fact? I've tried to work with lm_sensors,
> but the readings for that are way way off. So, considering  lm_sensors
> isuseless is there another way to tell if overheating is the problem?
>
> The case itself has a lot of fans, but it's also got 5 harddrives in it.
> On Tue, 6 Dec 2005, Daniel Gryniewicz wrote:
>
> > Date: Tue, 06 Dec 2005 18:39:48 -0500
> > From: Daniel Gryniewicz <dang@gentoo.org>
> > Reply-To: gentoo-amd64@lists.gentoo.org
> > To: gentoo-amd64@lists.gentoo.org
> > Subject: Re: [gentoo-amd64] mce log errors
> >
> > On Tue, 2005-12-06 at 14:56 -0600, Deedra Waters wrote:
> > > All,
> > >
> > > I'm getting a lot of these, but it only seems to happen when i put the
> > > machine under a lot of stress, and even then it's not always happening.
> > > This machine is a duel opteron 242, the board is an asus k8, and with
> > > the latest bios update, the machine has  no real problems at all.
> > >
> > > MCE 1
> > > CPU 0 4 northbridge TSC 8f1a7b270b6f
> > > ADDR 75c3320
> > >   Northbridge ECC error
> > >   ECC syndrome = 62
> > >        bit32 = err cpu0
> > >        bit46 = corrected ecc error
> > >   bus error 'local node origin, request didn't time out
> > >       generic read mem transaction
> > >       memory access, level generic'
> > > STATUS 9431400100000813 MCGSTATUS 0
> > > MCE 2
> > > CPU 0 2 bus unit TSC 8f8ad2325db7
> > >   L2 cache ECC error
> > >   Bus or cache array error
> > >        bit46 = corrected ecc error
> > >        bit62 = error overflow (multiple errors)
> > >   bus error 'local node origin, request didn't time out
> > >       prefetch mem transaction
> > >       memory access, level generic'
> > > STATUS d000400000000863 MCGSTATUS 0
> >
> > CPU cache is getting ECC errors.  Smells like overheating.
> >
> > Daniel
> >
>
>

-- 
Deedra Waters - Gentoo developer relations, accessibility and infrastructure -
dmwaters@gentoo.org
Gentoo linux: http://www.gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list

[gentoo-amd64] Re: mce log errors

[Duncan]

Deedra Waters posted <Pine.LNX.4.64.0512062202210.6176@monster>, excerpted
below,  on Tue, 06 Dec 2005 22:04:50 -0600:

> Is there a way to test that fact? I've tried to work with lm_sensors,
> but the readings for that are way way off. So, considering  lm_sensors
> isuseless is there another way to tell if overheating is the problem?
> 
> The case itself has a lot of fans, but it's also got 5 harddrives in it.

Don't know about ASUS, but Tyan has lm_sensors config files for many of
their boards on their site.

I had similar but not as severe (main memory only, no L2 cache errors)
here.  For quite some time they drove me nearly up a wall, so I can
definitely identify with your situation!

In my case, it turned out to be over-rated generic memory.  After a BIOS
update added memory timing control, I limited my so-called PC3200 memory
to PC3000 (downclocking from 200/400 MHz to 183/366 MHz), and now get to
actually enjoy that fabled Linux stability, with the only reboots being
when I do so purposefully!  =8^)

One thing you can try, somewhat counter-intuitive, but it definitely
helped here until BIOS got timing limit functionality (it didn't seem to
cause any compile problems or the like, either, a good thing on Gentoo),
is to turn OFF ECC. The best I can figure, the additional ECC data put a
higher strain on already touchy timings, so turning it OFF increased
stability while not noticably increasing undetected errors.

In any case, try declocking a bit.  I only declocked memory, but if it's
really L2 cache issues for you, you'll likely have to declock the CPUs as
well. If it's overheating or general timing touchiness, that should
definitely improve stability.

It could also be slightly low voltage.  Again, a properly configured
lm_sensors config would be a /great/ help here, but if it isn't
available...  Turning the clocking down should help there as well, but
turning the voltage up at the regular clock rate, provided your cooling is
fine, may also help.  If it's the cooling and NOT the voltage, that will
make things WORSE, of course, thereby giving you a way to tell the
difference, PROVIDED you want to risk 0v3r(10(kin9 methods even if not
actually overclocking, of course. =8^)

That leads to another possible solution, one some will certainly consider
more sane than resorting to upping voltages.  Particularly with that many
drives and the number of fans you indicate you may have, plus everything
else in a normal computer, it could be your power supply isn't quite large
enough to handle the demand.  This could easily be rehash for you, but
just in case...  many power supplies are hopelessly overrated,
particularly if you don't see any UL or CE (or other
appropriate nationality testing organization) certifications on them. If
they are certified, you can safely assume the rated overall output, but
particular voltages may still be inadequate for your needs, certainly so
when running five drives and possibly a fully loaded RAM rack, plus
multiiple fans.  Put it this way, if you are using the power supply that
came with the case, and you bought a low-end case, it's a fair bet that
the rating on the power supply isn't worth the sticker it's printed on!

Personally, I tend to prefer a pretty good power margin -- enough so it's
not even close to stressing, and I've never had the trouble with power
supplies I've seen others have.  If I'm spending an extra $50 to $100 to
have the peace of mind that stable power means, so be it!  Tt's well worth
it to me, considering the alternative of potential headaches and even
damage to a system worth conservatively a couple grand.

(I just need to learn to spend more on memory, as obviously, the generic
stuff I was buying didn't cut it.  <g>  Lesson learned!)

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html


-- 
gentoo-amd64@gentoo.org mailing list

[gentoo-amd64] Re: mce log errors

[Duncan]

Deedra Waters posted <Pine.LNX.4.64.0512062343090.6137@monster>, excerpted
below,  on Tue, 06 Dec 2005 23:44:40 -0600:

> Hrm, it doesn't look to be a heat issue. I opened the case and took a
> look inside when i saw the last message,and it looks perfectly cool and
> happy.
> 
> I do notice however, that it's only happening when i hammer the raid
> array, which is on a pci promise controler.

That would appear to be the voltage issue I mentioned in my last post,
that I was writing when you posted the above.  Pretty good odds, I'd say.

Also, since the RAID array would be powered directly off the power supply,
not from the board, that points squarely at it.  Replace it with a good
one.  Check ratings for the 12 volt rail hard drives usually use, in
particular.  Some of the higher end supplies actually have two
electrically separate 12-volt rails, distributing the stress.  However,
you probably don't actually /need/ that feature, as long as the rating for
the one is high enough to handle all those drives (plus additional
accessories, CD/DVD burner, perhaps, and USB bus powered devices and the
like).

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: mce log errors

[Deedra Waters]

Hrmmm,

The origional ps in this case was 500. We replaced it with 550 which
from what i gather should have been enough to handle the drives as well
as everything else, that's what i was told anyway. What would generally
be recommended for this much stuffed into a case like this?
On Tue, 6 Dec 2005, Duncan wrote:

> Date: Tue, 06 Dec 2005 23:11:00 -0700
> From: Duncan <1i5t5.duncan@cox.net>
> Reply-To: gentoo-amd64@lists.gentoo.org
> To: gentoo-amd64@lists.gentoo.org
> Subject: [gentoo-amd64]  Re: mce log errors
>
> Deedra Waters posted <Pine.LNX.4.64.0512062343090.6137@monster>, excerpted
> below,  on Tue, 06 Dec 2005 23:44:40 -0600:
>
> > Hrm, it doesn't look to be a heat issue. I opened the case and took a
> > look inside when i saw the last message,and it looks perfectly cool and
> > happy.
> >
> > I do notice however, that it's only happening when i hammer the raid
> > array, which is on a pci promise controler.
>
> That would appear to be the voltage issue I mentioned in my last post,
> that I was writing when you posted the above.  Pretty good odds, I'd say.
>
> Also, since the RAID array would be powered directly off the power supply,
> not from the board, that points squarely at it.  Replace it with a good
> one.  Check ratings for the 12 volt rail hard drives usually use, in
> particular.  Some of the higher end supplies actually have two
> electrically separate 12-volt rails, distributing the stress.  However,
> you probably don't actually /need/ that feature, as long as the rating for
> the one is high enough to handle all those drives (plus additional
> accessories, CD/DVD burner, perhaps, and USB bus powered devices and the
> like).
>
>

-- 
Deedra Waters - Gentoo developer relations, accessibility and infrastructure -
dmwaters@gentoo.org
Gentoo linux: http://www.gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: mce log errors

[Nuitari]

> The origional ps in this case was 500. We replaced it with 550 which
> from what i gather should have been enough to handle the drives as well
> as everything else, that's what i was told anyway. What would generally
> be recommended for this much stuffed into a case like this?

Which brand / model it is ?

You should be able to open the case and also see how much it is reader for 
each amp level that it does.
-- 
gentoo-amd64@gentoo.org mailing list

[gentoo-amd64] Re: Re: mce log errors

[Duncan]

Nuitari posted <Pine.LNX.4.64.0512070125180.23134@melchior.nuitari.net>,
excerpted below,  on Wed, 07 Dec 2005 01:26:03 -0500:

>> The origional ps in this case was 500. We replaced it with 550 which
>> from what i gather should have been enough to handle the drives as well
>> as everything else, that's what i was told anyway. What would generally
>> be recommended for this much stuffed into a case like this?
> 
> Which brand / model it is ?
> 
> You should be able to open the case and also see how much it is reader for
> each amp level that it does.

550 /should/ be enough, correct.  However, as Nuitari asks, what brand,
and is it UL/CE certified or not?  If not, as I said, the rating's worth
less than the sticker it's printed on.

Even the good brands have occasional duds.  I recently read about
someone having Enermax dud, for instance, and normally that's a decent
brand, from all I know.

You also mentioned dual Opteron 242, which is what I have.  My Tyan board
required a second connector for the additional power, an 8-pin "EPS-12V"
connector. It also has a 6-pin "SSI Workstation" connector for AGP-Pro,
but I didn't need that since I'm not running that powerful a video card. 
However, I expect I'd have issues if I didn't have the 8-pin plugged in. 
I had to buy a special power supply to get that.

Since we are both running similar equipment, different mobo but both
running dual Opteron 242, with RAID (mine is four-drive Seagate SATA 300
gig, plus an older 250 gig Maxtor PATA not in the RAID, plus I've both a
CD burner and a DVD burner), and I've had no problems with power here,
perhaps my power supply would be of interest?  (This assumes you have
the same 8-pin EPS-12v board connector and don't need the 6-pin AGP-Pro
thing.)

Vantec Stealth VAN-520A (520W sustained, 620W 60 second peak rating).

I purchased it at Fry's Electronics, tho similar rated supplies are
available online for somewhat cheaper.  (I like pricewatch.com, as I've
had decent experiences with their vendors, but even if you don't buy from
pricewatch listed vendors, it's a very quick way to come up with "street
price" estimates.  Even for stuff like power supplies and cases that I
usually buy locally, I'll check pricewatch to get an idea of the internet
price, before doing my local shopping.  Generally speaking, a price within
20% higher, more or less depending on item, will then be a decent
local brick and mortar sourced deal.)

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: mce log errors

[Nuitari]

> 550 /should/ be enough, correct.  However, as Nuitari asks, what brand,
> and is it UL/CE certified or not?  If not, as I said, the rating's worth
> less than the sticker it's printed on.

550w doesn't mean much actually. It is the total power output it is 
capable of, under ideal conditions.

You have to see how much each individual voltage can take amp wise.

In the time of 300w power supplies, there were a lot of cheap ones that 
had a 10A 12v line. 10a * 10v = 120watt, which is definitely not what 
someone would expect from a 300w power supply.

Then came the 350w ones with the same rating for 12v, so no gain there. 
Better ones had at least 15A for 12V as the CPUs started using much more 
power.

I am currently using an Antec SmartPower 2.0 450W power supply.

It has these ratings:
+3.3V * 32A  = 105.6W
+5V   * 30A  = 150W
+12V1 * 15A  = 180W
+12V2 * 17A  = 204W

According to the documentation, the combined output of these cannot exceed 
440Watts (there are some watts for +5V standby and -12V).

+12V1 is what does to the motherboard and peripherals and +12V2 goes to 
the processor(s).

You should calculate how much you are using power wise with all of your 
peripherals to see if your power supply is strong enough.
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: mce log errors

[Deedra Waters]

I don't know what brand, i'm totally blind, so couldn't read anything on
it to tell you. I do know that this board did require the 8 pin thing
for the board as yours does, and we did have to get a specific power
supply for that.
On Wed, 7 Dec 2005, Duncan wrote:

> Date: Wed, 07 Dec 2005 00:17:49 -0700
> From: Duncan <1i5t5.duncan@cox.net>
> Reply-To: gentoo-amd64@lists.gentoo.org
> To: gentoo-amd64@lists.gentoo.org
> Subject: [gentoo-amd64]  Re: Re: mce log errors
>
> Nuitari posted <Pine.LNX.4.64.0512070125180.23134@melchior.nuitari.net>,
> excerpted below,  on Wed, 07 Dec 2005 01:26:03 -0500:
>
> >> The origional ps in this case was 500. We replaced it with 550 which
> >> from what i gather should have been enough to handle the drives as well
> >> as everything else, that's what i was told anyway. What would generally
> >> be recommended for this much stuffed into a case like this?
> >
> > Which brand / model it is ?
> >
> > You should be able to open the case and also see how much it is reader for
> > each amp level that it does.
>
> 550 /should/ be enough, correct.  However, as Nuitari asks, what brand,
> and is it UL/CE certified or not?  If not, as I said, the rating's worth
> less than the sticker it's printed on.
>
> Even the good brands have occasional duds.  I recently read about
> someone having Enermax dud, for instance, and normally that's a decent
> brand, from all I know.
>
> You also mentioned dual Opteron 242, which is what I have.  My Tyan board
> required a second connector for the additional power, an 8-pin "EPS-12V"
> connector. It also has a 6-pin "SSI Workstation" connector for AGP-Pro,
> but I didn't need that since I'm not running that powerful a video card.
> However, I expect I'd have issues if I didn't have the 8-pin plugged in.
> I had to buy a special power supply to get that.
>
> Since we are both running similar equipment, different mobo but both
> running dual Opteron 242, with RAID (mine is four-drive Seagate SATA 300
> gig, plus an older 250 gig Maxtor PATA not in the RAID, plus I've both a
> CD burner and a DVD burner), and I've had no problems with power here,
> perhaps my power supply would be of interest?  (This assumes you have
> the same 8-pin EPS-12v board connector and don't need the 6-pin AGP-Pro
> thing.)
>
> Vantec Stealth VAN-520A (520W sustained, 620W 60 second peak rating).
>
> I purchased it at Fry's Electronics, tho similar rated supplies are
> available online for somewhat cheaper.  (I like pricewatch.com, as I've
> had decent experiences with their vendors, but even if you don't buy from
> pricewatch listed vendors, it's a very quick way to come up with "street
> price" estimates.  Even for stuff like power supplies and cases that I
> usually buy locally, I'll check pricewatch to get an idea of the internet
> price, before doing my local shopping.  Generally speaking, a price within
> 20% higher, more or less depending on item, will then be a decent
> local brick and mortar sourced deal.)
>
>

-- 
Deedra Waters - Gentoo developer relations, accessibility and infrastructure -
dmwaters@gentoo.org
Gentoo linux: http://www.gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list

[gentoo-amd64] Re: Re: Re: mce log errors

[Duncan]

Deedra Waters posted <Pine.LNX.4.64.0512070835120.6330@monster>, excerpted
below,  on Wed, 07 Dec 2005 08:35:30 -0600:

> I don't know what brand, i'm totally blind, so couldn't read anything on
> it to tell you. 

Wow!  I didn't know that!  I had seen a picture of you from one of the GWN
articles (the one covering the live bug day gathering up there in
Portland, last winter, IIRC), but didn't know /that/.  Cool!

That means you probably have a very good appreciation for the
Massachusetts OASIS standards debate, which MS turned to a question of
disability coverage.  I'd love to read it, but of course it's off topic
for here, unfortunately.  Still, if you have a blog or the like with your
thoughts on FLOSS accessibility both there and in general, feel free to
post a link!

They say the internet is a great equalizer, as it really doesn't matter
who you are, what you do, or what personal challenges you may have, only
what you know and are willing to share.  I have certainly found that to be
the case, here.

> I do know that this board did require the 8 pin thing for the board as
> yours does, and we did have to get a specific power supply for that.

It may be, then, that your supply is just a bad one, or that it's the CPU
itself that's bad.  I don't believe products with that 8-pin thing are
common enough to reach the low end in quality, so it's probably a one-off
deal, or as Nuitari suggests, simply a capacity issue on one of the
outputs while the others are fine.

FWIW, the ratings on my Vantec, from the manual, in the same order as
Nuitari's, for comparison, are:

+3.3V * 26A = 85.8W
+5V   * 52A = 260W
+12V  * 28A = 336W
(Only a single +12V rail)

The documentation limits the total of the +3.3 and +5V wattages also to
260W, the same as the +5V alone.  There's a total limit on all three, the
+3.3, +5, and +12 volt wattages, of 500 watt.

So... my overall rating is higher, as is my +5V rating.  +3.3V is a bit
lower, and I have only the single +12V rail, rated a bit lower than the
combined wattage of his two +12V rails.

As with him, there's also lower wattages available for +5VSB (stand-by),
-5V, and -12V.

If you can find the brand and model number of your supply, the numbers for
it will hopefully be available on the manufacturer's site, in a usable
format.

For point of power reference in relation to drive usage, my Seagate SATAs
say maximum spin-up draw is 2.8 amp @ 12 volt, so roughly 11 amp if all
four are spun up at once, 14 if the fifth drive is assumed to have
similar power requirements. Operational draw is of course much lower, .85
amp on the 5 volt, .79 amp on the 12 volt, according to specs, so 4.3 amp
5 volt, 4 amp 12 volt draw, five disk operational.  Say 5 amp total
for each voltage, 5 and 12.

Of course, to that you must add memory/CPU/mobo/fan/expansion use. 
However, 25 amps on the 12 volt should be plenty, unless you have
unusually high draw expansion needs.  It /may/ be your 5 volt giving out,
given the disks use that as well in operational mode, and 5 volt use is
more common operational use on bus/mobo powered peripherals and expansion
cards, as well.  Note how high the 5 volt amp rating on mine is, and I'm
not having issues, with a setup fairly similar to yours.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: Re: mce log errors

[Homer Parker]

On Wed, 2005-12-07 at 10:45 -0700, Duncan wrote:
> Of course, to that you must add memory/CPU/mobo/fan/expansion use. 
> However, 25 amps on the 12 volt should be plenty, unless you have
> unusually high draw expansion needs.  It /may/ be your 5 volt giving
> out,
> given the disks use that as well in operational mode, and 5 volt use
> is
> more common operational use on bus/mobo powered peripherals and
> expansion
> cards, as well.  Note how high the 5 volt amp rating on mine is, and
> I'm
> not having issues, with a setup fairly similar to yours.

	The specs on the PSU in her system are:

Total Continuous Output 550W
+5V 40A
+12V1 19A
+12V2 19A
+3.3V 32A
+5Vsb 2.0A

	I'm still working up total consumption, but so far it's way under those
specs.

-- 
Homer Parker
Gentoo/AMD64 Team
Gentoo/AMD64 Arch Tester Strategic Lead
Gentoo Linux Developer Relations
hparker@gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: Re: mce log errors

[Billy Holmes]

Duncan wrote:
> for here, unfortunately.  Still, if you have a blog or the like with your
> thoughts on FLOSS accessibility both there and in general, feel free to
> post a link!

Aye, I'd be mega-interested in reading that as well.
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: Re: mce log errors

[Nuitari]

> For point of power reference in relation to drive usage, my Seagate SATAs
> say maximum spin-up draw is 2.8 amp @ 12 volt, so roughly 11 amp if all
> four are spun up at once, 14 if the fifth drive is assumed to have
> similar power requirements. Operational draw is of course much lower, .85
> amp on the 5 volt, .79 amp on the 12 volt, according to specs, so 4.3 amp
> 5 volt, 4 amp 12 volt draw, five disk operational.  Say 5 amp total
> for each voltage, 5 and 12.

Yes, although that a lot of factory default settings for scsi hard drives 
have a Delayed start option of 10s * scsi ID and / or until the start 
command is received so that the power supply doesn't get hammered.
-- 
gentoo-amd64@gentoo.org mailing list

[gentoo-amd64] Re: Re: Re: Re: mce log errors

[Duncan]

Nuitari posted <Pine.LNX.4.64.0512071333150.23032@melchior.nuitari.net>,
excerpted below,  on Wed, 07 Dec 2005 13:36:39 -0500:

> Yes, although that a lot of factory default settings for scsi hard drives
> have a Delayed start option of 10s * scsi ID and / or until the start
> command is received so that the power supply doesn't get hammered.

Indeed.  I was giving that as a worst-case, however, and in any case, I'm
not sure SATA, which is what I'm using (don't know about her, she says
PCI Promise controller, which hints SATA to me, but doesn't specify), does
that delayed stuff. Does anyone know if it does? I'd assume not, otherwise.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html


-- 
gentoo-amd64@gentoo.org mailing list

[gentoo-amd64] Re: Re: Re: Re: mce log errors

[Duncan]

Homer Parker posted <1133979787.10108.17.camel@laptop>, excerpted below, 
on Wed, 07 Dec 2005 12:23:07 -0600:

> 	The specs on the PSU in her system are:
> 
> Total Continuous Output 550W
> +5V 40A
> +12V1 19A
> +12V2 19A
> +3.3V 32A
> +5Vsb 2.0A
> 
> 	I'm still working up total consumption, but so far it's way under those
> specs.

Yeah.  Nearly 40 amp total +12 volt, 40 amp 5 volt, that should be a
plenty for nearly anything (as Mr. Gates is alleged to have said about
640k <g>).

Thus, it's looking to me like something, probably the power supply, is
defective.

Just thinking...  It isn't bad location power, one would hope?  A power
supply can only do so much, if the input power is bad.  I know we're about
to move out of an old building at work, which gives the computers fits
from time to time, even with them on (little/local, not building) UPSs to
filter the power.  (Opening day is scheduled two weeks from today, the
21st!  Whee!  Christmas present!  It means I don't get to leave for the
holidays, tho.)

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: Re: Re: mce log errors

[Deedra Waters]

Yes, sata. 4 sata drives, and a pata drive, plus dvdrw.
On Wed, 7 Dec 2005, Duncan wrote:

> Date: Wed, 07 Dec 2005 12:50:30 -0700
> From: Duncan <1i5t5.duncan@cox.net>
> Reply-To: gentoo-amd64@lists.gentoo.org
> To: gentoo-amd64@lists.gentoo.org
> Subject: [gentoo-amd64]  Re: Re: Re: Re: mce log errors
>
> Nuitari posted <Pine.LNX.4.64.0512071333150.23032@melchior.nuitari.net>,
> excerpted below,  on Wed, 07 Dec 2005 13:36:39 -0500:
>
> > Yes, although that a lot of factory default settings for scsi hard drives
> > have a Delayed start option of 10s * scsi ID and / or until the start
> > command is received so that the power supply doesn't get hammered.
>
> Indeed.  I was giving that as a worst-case, however, and in any case, I'm
> not sure SATA, which is what I'm using (don't know about her, she says
> PCI Promise controller, which hints SATA to me, but doesn't specify), does
> that delayed stuff. Does anyone know if it does? I'd assume not, otherwise.
>
>

-- 
Deedra Waters - Gentoo developer relations, accessibility and infrastructure -
dmwaters@gentoo.org
Gentoo linux: http://www.gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list

[gentoo-amd64] Re: Re: Re: Re: Re: mce log errors

[Duncan]

Deedra Waters posted <Pine.LNX.4.64.0512071409440.6178@monster>, excerpted
below,  on Wed, 07 Dec 2005 14:10:07 -0600:

> Yes, sata. 4 sata drives, and a pata drive, plus dvdrw.

<g>  Your system is so much like mine, it's downright eerie!  Great mind
think alike, they say!  =8^)

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: Re: Re: mce log errors

[Deedra Waters]

Power at this house is fine. We've not had any problems as far as i can
tell, and the box is on a ups that's good, so i don't think it's that
either.

So, no overheating, I seriously dout the ps is defective, though i guess
anything is possible. What else would that leave? heh
On Wed, 7 Dec 2005, Duncan wrote:

> Date: Wed, 07 Dec 2005 13:01:50 -0700
> From: Duncan <1i5t5.duncan@cox.net>
> Reply-To: gentoo-amd64@lists.gentoo.org
> To: gentoo-amd64@lists.gentoo.org
> Subject: [gentoo-amd64]  Re: Re: Re: Re: mce log errors
>
> Homer Parker posted <1133979787.10108.17.camel@laptop>, excerpted below,
> on Wed, 07 Dec 2005 12:23:07 -0600:
>
> > 	The specs on the PSU in her system are:
> >
> > Total Continuous Output 550W
> > +5V 40A
> > +12V1 19A
> > +12V2 19A
> > +3.3V 32A
> > +5Vsb 2.0A
> >
> > 	I'm still working up total consumption, but so far it's way under those
> > specs.
>
> Yeah.  Nearly 40 amp total +12 volt, 40 amp 5 volt, that should be a
> plenty for nearly anything (as Mr. Gates is alleged to have said about
> 640k <g>).
>
> Thus, it's looking to me like something, probably the power supply, is
> defective.
>
> Just thinking...  It isn't bad location power, one would hope?  A power
> supply can only do so much, if the input power is bad.  I know we're about
> to move out of an old building at work, which gives the computers fits
> from time to time, even with them on (little/local, not building) UPSs to
> filter the power.  (Opening day is scheduled two weeks from today, the
> 21st!  Whee!  Christmas present!  It means I don't get to leave for the
> holidays, tho.)
>
>

-- 
Deedra Waters - Gentoo developer relations, accessibility and infrastructure -
dmwaters@gentoo.org
Gentoo linux: http://www.gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: Re: Re: mce log errors

[Brett Johnson]

I tend to agree with the power supply theory. Was the make and model of
the PS ever determined?

In the Jan 2005 issue of Maximum PC
(http://www.maximumpc.com/issue_date/january_2005/) there is an atricle
testing 7 ~500 watt power supplies and a 400 watt reference power supply.

Of the 8 PS tested, 5 including the referenced failed the tests. One power
supply rated at 520 watts could not even sustain 284 watts of output
(drawing 378 watts of AC power multiplied by an efficiency factor of 75%
is roughly 284 watts). It repeatedly crashed the test system during the
load tests. Another 500 watt power supply did not crash the system, but
caused strange white lines to appear on a black test screen and emitted a
burning smell.

The test system was a 3.4 GHz P4 with 3GB ram, GeFroce 6800 256MB video
card, 2 7,200rpm ata drives and 2 15,000rpm scsi drives, 2 optical drives,
SB Live! sound card, NIC and 2 120mm case fans.

The results: not all power supplies are equal, and strange things can
happen when the system is starved of power.

Brett

> Power at this house is fine. We've not had any problems as far as i can
> tell, and the box is on a ups that's good, so i don't think it's that
> either.
>
> So, no overheating, I seriously dout the ps is defective, though i guess
> anything is possible. What else would that leave? heh
> On Wed, 7 Dec 2005, Duncan wrote:
>



Random Thought:
---------------
* m2 stares at the monitor... it looks like a hamburger...
<Knghtbrd> m2 - that's a bad sign

-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: Re: Re: mce log errors

[Deedra Waters]

All,

Thanks a great deal for all who helped. I had the PS tested, it came
back good. However, apparrently the ups i had it hooked up to wasn't
putting out enough juice for the thing it seems. New huge ups, and the
machine appears  to be happy.:)
On Wed, 7 Dec 2005, Brett Johnson wrote:

> Date: Wed, 7 Dec 2005 15:49:50 -0600 (CST)
> From: Brett Johnson <brett@blzj.com>
> Reply-To: gentoo-amd64@lists.gentoo.org
> To: gentoo-amd64@lists.gentoo.org
> Subject: Re: [gentoo-amd64]  Re: Re: Re: Re: mce log errors
>
> I tend to agree with the power supply theory. Was the make and model of
> the PS ever determined?
>
> In the Jan 2005 issue of Maximum PC
> (http://www.maximumpc.com/issue_date/january_2005/) there is an atricle
> testing 7 ~500 watt power supplies and a 400 watt reference power supply.
>
> Of the 8 PS tested, 5 including the referenced failed the tests. One power
> supply rated at 520 watts could not even sustain 284 watts of output
> (drawing 378 watts of AC power multiplied by an efficiency factor of 75%
> is roughly 284 watts). It repeatedly crashed the test system during the
> load tests. Another 500 watt power supply did not crash the system, but
> caused strange white lines to appear on a black test screen and emitted a
> burning smell.
>
> The test system was a 3.4 GHz P4 with 3GB ram, GeFroce 6800 256MB video
> card, 2 7,200rpm ata drives and 2 15,000rpm scsi drives, 2 optical drives,
> SB Live! sound card, NIC and 2 120mm case fans.
>
> The results: not all power supplies are equal, and strange things can
> happen when the system is starved of power.
>
> Brett
>
> > Power at this house is fine. We've not had any problems as far as i can
> > tell, and the box is on a ups that's good, so i don't think it's that
> > either.
> >
> > So, no overheating, I seriously dout the ps is defective, though i guess
> > anything is possible. What else would that leave? heh
> > On Wed, 7 Dec 2005, Duncan wrote:
> >
>
>
>
> Random Thought:
> ---------------
> * m2 stares at the monitor... it looks like a hamburger...
> <Knghtbrd> m2 - that's a bad sign
>
>

-- 
Deedra Waters - Gentoo developer relations, accessibility and infrastructure -
dmwaters@gentoo.org
Gentoo linux: http://www.gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list

[gentoo-amd64] Re: Re: Re: Re: Re: mce log errors

[Duncan]

Deedra Waters posted <Pine.LNX.4.64.0512102119300.6755@monster>, excerpted
below,  on Sat, 10 Dec 2005 21:23:22 -0600:

> Thanks a great deal for all who helped. I had the PS tested, it came
> back good. However, apparrently the ups i had it hooked up to wasn't
> putting out enough juice for the thing it seems. New huge ups, and the
> machine appears  to be happy.:)

Very cool you got it working!  Of course the "PS" in UPS is "Power
Supply", so it was the power supply after all, just not the one we all
thought!  =8^)

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html


-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: Re: Re: Re: mce log errors

[Nuitari]

>> Thanks a great deal for all who helped. I had the PS tested, it came
>> back good. However, apparrently the ups i had it hooked up to wasn't
>> putting out enough juice for the thing it seems. New huge ups, and the
>> machine appears  to be happy.:)

Do you have the brand / modewl of the faulty UPS ?
-- 
gentoo-amd64@gentoo.org mailing list

Re: [gentoo-amd64] Re: Re: Re: Re: Re: mce log errors

[Deedra Waters]

UPS wasn't faulty, it was just too small. 300   wats ups, when the ps for this
machine is 550 wats. I didn't think about the too small factor for the
ups till the thing started turning itself off 2 nights ago:p

On Sun, 11 Dec 2005, Nuitari wrote:

> Date: Sun, 11 Dec 2005 12:15:44 -0500 (EST)
> From: Nuitari <nuitari@melchior.nuitari.net>
> Reply-To: gentoo-amd64@lists.gentoo.org
> To: gentoo-amd64@lists.gentoo.org
> Subject: Re: [gentoo-amd64]  Re: Re: Re: Re: Re: mce log errors
>
> > > Thanks a great deal for all who helped. I had the PS tested, it came
> > > back good. However, apparrently the ups i had it hooked up to wasn't
> > > putting out enough juice for the thing it seems. New huge ups, and the
> > > machine appears  to be happy.:)
>
> Do you have the brand / modewl of the faulty UPS ?
>

-- 
Deedra Waters - Gentoo developer relations, accessibility and infrastructure -
dmwaters@gentoo.org
Gentoo linux: http://www.gentoo.org

-- 
gentoo-amd64@gentoo.org mailing list