6:54 p.m.
I have a most distressing problem. I am a faculty member in the computer
and information sciences department at a liberal arts college.
We have a lab with a number of computers with two AMD 32 bit CPUs and a few
computers with two AMD 64 bit CPUs.
Two years ago we ran FC4 on both types of these computers with no problem.
However, attempting to install FC5 has killed several computers. We have
now burned out two of the 64 bit computers and two of the 32 bit computers.
Generally, the behavior is that they go a ways into the installation process
and hang. After retrying the installation another time, the computer is
basically dead; when I tried a Knoppix CD-ROM one of them could not get past
the screen telling you to press a key to boot Knoppix. One of the 64 bit
computers went "poof", and I smelled smoke. It is,of course, completely
dead.
I've used Google to see if I could find out about a problem to no avail.
We're getting pretty desperate. Can anyone offer any guidance? At this
point, we're even afraid to try another distribution. We're starting to
install Windows to get the remaining computers working.
I'll be glad to provide more details if they will help.
Thank you in advance..
7:25 p.m.
On Fri, 2006-08-18 at 19:54 -0400, Robert Gann wrote:
I have a most distressing problem. I am a faculty member in the
computer and information sciences department at a liberal arts
college.
We have a lab with a number of computers with two AMD 32 bit CPUs and
a few computers with two AMD 64 bit CPUs.
Two years ago we ran FC4 on both types of these computers with no
problem. However, attempting to install FC5 has killed several
computers. We have now burned out two of the 64 bit computers and two
of the 32 bit computers. Generally, the behavior is that they go a
ways into the installation process and hang. After retrying the
installation another time, the computer is basically dead; when I
tried a Knoppix CD-ROM one of them could not get past the screen
telling you to press a key to boot Knoppix. One of the 64 bit
computers went “poof”, and I smelled smoke. It is,of course,
completely dead.
I’ve used Google to see if I could find out about a problem to no
avail. We’re getting pretty desperate. Can anyone offer any
guidance? At this point, we’re even afraid to try another
distribution. We’re starting to install Windows to get the remaining
computers working.
I’ll be glad to provide more details if they will help.
Thank you in advance..
Robert,
First and foremost, there is not way an operating system, -any-
operating could burn a PC; let alone the OS's installer. Period.
The only way, I'm aware of, to damage a system using software is to
slow-down the CPU/MB fans using special hardware/i2c control software.
AFAIK, there's no such software in Fedora Core.
(There's a known issue that might cause Linux to trash old-CDROM's
firmware but this certainly cannot cause a machine to burn).
In general, smoke that is followed by computer death usually means:
A. Damaged PSU (power supply unit).
B. Power surge.
C. Over-heating. (Dead CPU/MB fan. Though new computer should shut down
in-case of over heating.)
D. Short circuit in one of the computer components, most likely the
mother board.
Reading through your description of the problem I'd wager it's 3.
Open the machine's case, check the CPU and mother board FANs. Are they
spinning? Do you see large piles of dust on the heat sinks? (Large
metallic object under the fan). Is the heat sink seated firmly?
As I said. This is purely a hardware problem. Installing Windows will
only kill more machines.
Gilboa
7:42 p.m.
On Sat, 19 Aug 2006, Gilboa Davara wrote:
[...]
In general, smoke that is followed by computer death usually means:
A. Damaged PSU (power supply unit).
B. Power surge.
C. Over-heating. (Dead CPU/MB fan. Though new computer should shut down
in-case of over heating.)
D. Short circuit in one of the computer components, most likely the
mother board.
Additionally, I have seen peripherals burn out motherboards. My boss
toasted two machines by plugging a bad printer into them successively. If
you are burning out multiple machines the first question is 'what is
common'. But the OS specifically is extremely unlikely to be the cause of
multiple machine burnouts.
Bad power (all of your computers are plugged into good quality UPS units
and your building power and grounds are good, right?), bad attached
devices (PCI cards, USB drives, printers, monitors, even keyboards (think
'soda in keyboard') and mice are possible), 'toxic environment' (if your
office is 110 degrees or in a damp basement you may have your problem
right there), or unknown damage while in storage (IOW: Maybe someone
stored them in a toxic environment while they were turned off and got them
wet or something...).
I concur with the below that installing windows instead will probably only
result in more dead machines until you identify your hardware or
environmental issue.
Reading through your description of the problem I'd wager
it's 3.
Open the machine's case, check the CPU and mother board FANs. Are they
spinning? Do you see large piles of dust on the heat sinks? (Large
metallic object under the fan). Is the heat sink seated firmly?
As I said. This is purely a hardware problem. Installing Windows will
only kill more machines.
Gilboa
--
Benjamin Franz
"It is moronic to predict without first establishing an error rate
for a prediction and keeping track of one’s past record of accuracy."
-- Nassim Nicholas Taleb, Fooled By Randomness
8:49 p.m.
Test the power supplies. Buy a $20 power supply tester and plug the 20 or 24 pin power
connector into it. Smoke is telling you there is a physical fault with the computer, not a
software one.
Bob Cochran
----- Original Message -----
From: Robert Gann
To: fedora-list(a)redhat.com
Sent: Friday, August 18, 2006 7:54 PM
Subject: FC5 Installation Killing Computers
I have a most distressing problem. I am a faculty member in the computer and
information sciences department at a liberal arts college.
We have a lab with a number of computers with two AMD 32 bit CPUs and a few computers
with two AMD 64 bit CPUs.
Two years ago we ran FC4 on both types of these computers with no problem. However,
attempting to install FC5 has killed several computers. We have now burned out two of the
64 bit computers and two of the 32 bit computers. Generally, the behavior is that they go
a ways into the installation process and hang. After retrying the installation another
time, the computer is basically dead; when I tried a Knoppix CD-ROM one of them could not
get past the screen telling you to press a key to boot Knoppix. One of the 64 bit
computers went "poof", and I smelled smoke. It is,of course, completely dead.
I've used Google to see if I could find out about a problem to no avail. We're
getting pretty desperate. Can anyone offer any guidance? At this point, we're even
afraid to try another distribution. We're starting to install Windows to get the
remaining computers working.
I'll be glad to provide more details if they will help.
Thank you in advance..
------------------------------------------------------------------------------
--
fedora-list mailing list
fedora-list(a)redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
10:38 p.m.
From: Robert Gann
To: fedora-list(a)redhat.com
Sent: Friday, August 18, 2006 7:54 PM
Subject: FC5 Installation Killing Computers
I have a most distressing problem. I am a faculty member in
the computer and information sciences department at a liberal
arts college.
...and the first step to recovery is to admit that you have a problem.
<cackles> Ric
--
================================================
My father, Victor Moore (Vic) used to say:
"There are two Great Sins in the world...
...the Sin of Ignorance, and
...the Sin of Stupidity.
Only the former may be overcome." R.I.P. Dad.
Linux user# 44256
Sign up at: http://counter.li.org/
http://www.sourceforge.net/projects/oar
http://www.wayward4now.net
================================================
10:27 p.m.
On Fri, 2006-08-18 at 19:54 -0400, Robert Gann wrote:
Two years ago we ran FC4 on both types of these computers with no
problem. However, attempting to install FC5 has killed several
computers. We have now burned out two of the 64 bit computers and two
of the 32 bit computers. Generally, the behavior is that they go a
ways into the installation process and hang. After retrying the
installation another time, the computer is basically dead; when I
tried a Knoppix CD-ROM one of them could not get past the screen
telling you to press a key to boot Knoppix. One of the 64 bit
computers went “poof”, and I smelled smoke. It is,of course,
completely dead.
Did you take the computers to the vendor you bought them from? It's not
the OS, it's more likely to be power supplies, which burn out right and
left lately. Mine went ape on me, it took about 5 days for the power
supply to blow, which pinpointed the problem. Ric
--
================================================
My father, Victor Moore (Vic) used to say:
"There are two Great Sins in the world...
...the Sin of Ignorance, and
...the Sin of Stupidity.
Only the former may be overcome." R.I.P. Dad.
Linux user# 44256
Sign up at: http://counter.li.org/
http://www.sourceforge.net/projects/oar
http://www.wayward4now.net
================================================
12:24 p.m.
Robert Gann wrote:
We have a lab with a number of computers with two AMD 32 bit CPUs and a
few computers with two AMD 64 bit CPUs.
Two years ago we ran FC4 on both types of these computers with no
problem.
Have these computer you are having problems with run Fedora before? Or
Linux before?
However, attempting to install FC5 has killed several
computers. We have now burned out two of the 64 bit computers and
two
of the 32 bit computers. Generally, the behavior is that they go a ways
into the installation process and hang. After retrying the installation
another time, the computer is basically dead; when I tried a Knoppix
CD-ROM one of them could not get past the screen telling you to press a
key to boot Knoppix. One of the 64 bit computers went “poof”, and I
smelled smoke. It is,of course, completely dead.
I have had this happen due to hardware problems. Heck, even burning a
CD was causing me problems due to a power supply that was just at it's
limit.
Depending on the hardware, running Linux on it can push cheap hardware
past it's limits. Power supplies that are rated XXX watts but this is
actully peak power.
Multitasking/mutithreading can load a system more than just the old
windows way of doing things. This causes the processors and power
supplies to work harder. Dust in the cases or as someone else
suggested, slow or non-working fans on top of the extra work has caused
things to overheat.
FC5 won't be burning the computers out but it could be showing the
weaknesses of the hardware.
--
Robin Laing
1:10 p.m.
I dunno...I saw another thread here stating that FC5 was killing HDs by
restarting them repeatedly until they died so this does seem like it could
be possible...makes me glad I'm not thinking of using FC5 nor currently
using it and such a distro which has now had multiple posts with regards to
it killing various computer components certainly is not good for this
distros reputation nor Linuxes as a whole, the developers of this distro
might want to make fixing these issues a rather high priority if they care
about FCs reputation at all...
On 8/21/06, Robin Laing <Robin.Laing(a)drdc-rddc.gc.ca> wrote:
Robert Gann wrote:
>
> We have a lab with a number of computers with two AMD 32 bit CPUs and a
> few computers with two AMD 64 bit CPUs.
>
>
>
> Two years ago we ran FC4 on both types of these computers with no
> problem.
Have these computer you are having problems with run Fedora before? Or
Linux before?
However, attempting to install FC5 has killed several
> computers. We have now burned out two of the 64 bit computers and two
> of the 32 bit computers. Generally, the behavior is that they go a ways
> into the installation process and hang. After retrying the installation
> another time, the computer is basically dead; when I tried a Knoppix
> CD-ROM one of them could not get past the screen telling you to press a
> key to boot Knoppix. One of the 64 bit computers went "poof", and I
> smelled smoke. It is,of course, completely dead.
>
I have had this happen due to hardware problems. Heck, even burning a
CD was causing me problems due to a power supply that was just at it's
limit.
Depending on the hardware, running Linux on it can push cheap hardware
past it's limits. Power supplies that are rated XXX watts but this is
actully peak power.
Multitasking/mutithreading can load a system more than just the old
windows way of doing things. This causes the processors and power
supplies to work harder. Dust in the cases or as someone else
suggested, slow or non-working fans on top of the extra work has caused
things to overheat.
FC5 won't be burning the computers out but it could be showing the
weaknesses of the hardware.
--
Robin Laing
--
fedora-list mailing list
fedora-list(a)redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
1:24 p.m.
Hex Star wrote:
I dunno...I saw another thread here stating that FC5 was killing HDs
by
restarting them repeatedly until they died so this does seem like it
could be possible...makes me glad I'm not thinking of using FC5 nor
currently using it and such a distro which has now had multiple posts
with regards to it killing various computer components certainly is not
good for this distros reputation nor Linuxes as a whole, the developers
of this distro might want to make fixing these issues a rather high
priority if they care about FCs reputation at all...
Aaaaahh hold on I had two HDDs die that way on one box here, I don't
know what caused it but the most likely explanation is the powersupply
in the box browning out. The fact there was only one other potential
sufferer I could find suggests that it is not the OS but the hardware.
So please don't take away from my experience that there is anything
wrong with FC5 on the whole, it is in use on several machines here for
many months with no problems like that and I can and do recommend it and
stick it on others' machines.
-Andy
9:40 p.m.
On Monday 21 August 2006 14:10, Hex Star wrote:
I dunno...I saw another thread here stating that FC5 was killing
HDs by restarting them repeatedly until they died so this does seem
like it could be possible...makes me glad I'm not thinking of using
FC5 nor currently using it and such a distro which has now had
multiple posts with regards to it killing various computer
components certainly is not good for this distros reputation nor
Linuxes as a whole, the developers of this distro might want to
make fixing these issues a rather high priority if they care about
FCs reputation at all...
You are taking things out of context. Install a copy on a box and
check it out for yourself. Out of personal preference I choose not
to use none RPM flavored distros of Linux but I don't attack them
because of my choice not to use them.
You people that for whatever reason decide that a distro is bad for
whatever reason like to find anything you can to bad mouth it. How
is attacking a Linux distro accomplishing anything good?
If you don't use FC and don't want too... why are you on the mailing
list? Try finding facts and stop spreading FUD!
--
Jack Gates http://www.morningstarcom.net
9:45 p.m.
No thanks, I won't be installing FC, at least until these issues are sorted
out as I value my PCs life ;-)
On 8/21/06, Jack Gates <jlgates(a)charter.net> wrote:
On Monday 21 August 2006 14:10, Hex Star wrote:
> I dunno...I saw another thread here stating that FC5 was killing
> HDs by restarting them repeatedly until they died so this does seem
> like it could be possible...makes me glad I'm not thinking of using
> FC5 nor currently using it and such a distro which has now had
> multiple posts with regards to it killing various computer
> components certainly is not good for this distros reputation nor
> Linuxes as a whole, the developers of this distro might want to
> make fixing these issues a rather high priority if they care about
> FCs reputation at all...
You are taking things out of context. Install a copy on a box and
check it out for yourself. Out of personal preference I choose not
to use none RPM flavored distros of Linux but I don't attack them
because of my choice not to use them.
You people that for whatever reason decide that a distro is bad for
whatever reason like to find anything you can to bad mouth it. How
is attacking a Linux distro accomplishing anything good?
If you don't use FC and don't want too... why are you on the mailing
list? Try finding facts and stop spreading FUD!
--
Jack Gates http://www.morningstarcom.net
--
fedora-list mailing list
fedora-list(a)redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
10:40 p.m.
On Mon, 2006-08-21 at 19:45 -0700, Hex Star wrote:
No thanks, I won't be installing FC, at least until these issues
are
sorted out as I value my PCs life ;-)
----
that of course is your prerogative.
I suspect that there are thousands if not hundreds of thousands of FC5
installs that were without any of the problems reported by that one
person.
More to the point, there is bugzilla and the proper place to problems
with Red Hat and Fedora Core software is bugzilla and I'm wondering if
you can reference a report on bugzilla that concerns this issue -
otherwise, you risk appearing to be the children's fable character
Chicken Little pronouncing the sky is falling when there are so many who
can that it is not.
You can't take untrusted users reports on this list seriously - they
have to make a proper bug report and bugzilla is the place - it's not a
bug if it isn't in bugzilla. The reporter that you are referring to did
not have a credible report.
Craig
1:40 a.m.
Craig White wrote:
You can't take untrusted users reports on this list seriously -
they
have to make a proper bug report and bugzilla is the place - it's not a
bug if it isn't in bugzilla. The reporter that you are referring to did
not have a credible report.
Noting a problem on the list and taking the temperature of the replies
is a useful tool. If five or six people had similar experiences one
could draw something very different from my situation. Sticking it in
Bugzilla is a different level of visibility. However, despite I
couldn't convince myself or anyone else it was an OS problem in the end,
I and my two busted HDDs assure you the report is 'credible' and not
caused by some hallucination.
I think Hex Star will find an interesting feature of support lists or
forums... they're full of problems. That's true for every OS. Some of
the problems are real and effect everyone, some affect just a few users
in special circumstances and always some problems affect just the
reporter :-)
-Andy
1:33 a.m.
Definitions of *cease* on the Web:
- lubricate with grease; "grease the wheels"
how do I lubricate posting? :-/
On 8/21/06, Chris Mohler <cr33dog(a)gmail.com> wrote:
On 8/21/06, Hex Star <hexstar(a)gmail.com> wrote:
> No thanks, I won't be installing FC, at least until these issues are
sorted
> out as I value my PCs life ;-)
Well, then cease posting.
Chris
--
fedora-list mailing list
fedora-list(a)redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
2:43 a.m.
Hex Star wrote:
Definitions of *cease* on the Web:
* lubricate with grease; "grease the wheels"
how do I lubricate posting? :-/
Ahhh.... I think you may have mistyped...or you're using a very bad
dictionary site.
http://www.webster.com/cgi-bin/dictionary?sourceid=Mozilla-search&va=...
Besides, a "lubricated posting" would probably fall under the heading of
Spam anyway.
--
Each person has the right to take the subway.
3:34 a.m.
Robin Laing:
> Depending on the hardware, running Linux on it can push cheap
hardware
> past it's limits.
>
> Multitasking/mutithreading can load a system more than just the old
> windows way of doing things. This causes the processors and power
> supplies to work harder.
I've read this before, but I find it hard to believe. I regularly see
Windows systems where the CPU maxes out a lot, doing ordinary tasks.
I've seen many where the resources are used almost to the full almost
all of the time. But on the same hardware using Linux, instead, I've
usually found that it spends most of its time being idle. Sure, you can
have some heavy intensive tasks, but that doesn't tend be the normal way
that Linux runs.
> FC5 won't be burning the computers out but it could be
showing the
> weaknesses of the hardware.
As could any strenuous task.
Hex Star:
I dunno...I saw another thread here stating that FC5 was killing HDs
by restarting them repeatedly until they died so this does seem like
it could be possible...
If a hard drive is failing, and sits there going "click whirr whump",
repeatedly, it's the hard drive that does that all by itself. You can
demonstrate this by powering up such a drive without any operating
system running.
If you're referring to a system that parks a drive when it's not being
used for a while (sleeping/hibernating/etc.), I've not seen Linux do
that by default, but it's a default behaviour for a few versions of
Windows.
makes me glad I'm not thinking of using FC5 nor currently using
it and
such a distro which has now had multiple posts with regards to it
killing various computer components certainly is not good for this
distros reputation nor Linuxes as a whole, the developers of this
distro might want to make fixing these issues a rather high priority
if they care about FCs reputation at all...
This thread has the hallmarks of troll FUDing (questionable claims,
dubious warnings, denigrations, scaremongering, etc.).
--
(Currently running FC4, in case that's important to the thread)
Don't send private replies to my address, the mailbox is ignored.
I read messages from the public lists.
4:46 a.m.
Tim wrote:
If a hard drive is failing, and sits there going "click whirr
whump",
repeatedly, it's the hard drive that does that all by itself. You can
demonstrate this by powering up such a drive without any operating
system running.
Good point. I just powered up (using a different PSU) both drives
disconnected from anything else for an hour, no clicks. Just sayin'.
If you're referring to a system that parks a drive when it's
not being
used for a while (sleeping/hibernating/etc.), I've not seen Linux do
that by default, but it's a default behaviour for a few versions of
Windows.
What was observed here by tailing /var/log/messages were log entries
along the lines of "SATA removed" then "SATA added" as if the drive
was
being rediscovered, happening at intervals of ten seconds or so IIRC.
-Andy
7:45 p.m.
On Tue, 2006-08-22 at 10:46 +0100, Andy Green wrote:
What was observed here by tailing /var/log/messages were log entries
along the lines of "SATA removed" then "SATA added" as if the drive
was
being rediscovered, happening at intervals of ten seconds or so IIRC.
Whoot! Dare it is! Whoot! Dare it is! Ric
6:28 a.m.
On Tue, 2006-08-22 at 18:04 +0930, Tim wrote:
This thread has the hallmarks of troll FUDing (questionable claims,
dubious warnings, denigrations, scaremongering, etc.).
I think that's a bit extreme. In cases like this, I find that it helps
to be a little patient because there's probably a little something else
(or combination of some things) that is happening that either the OP did
not notice or think was significant.
There is that well-known urban legend about the "cursed" hospital bed
where any patient who lies in it is dead by morning, and the staff
eventually found out that the cleaning crew would turn off the life
support machine every morning to use the electrical socket.
The story may not be true, but I have seen my fair share of strange
things that happen, which were initially attributed to something, when
in fact it was something else.
Let's all wait for the other shoe to drop.
--
Pascal Chong
email: chongym(a)cymulacrum.net
web: http://cymulacrum.net
pgp: http://cymulacrum.net/pgp/cymulacrum.asc
"La science ne connaît pas de frontière parce que la connaissance
appartient à l’humanité. et que c’est la flamme qui illumine le monde."
-- Louis Pasteur
7:04 a.m.
New subject: Crappy HDDs Killing Computers
Chong Yu Meng wrote:
Let's all wait for the other shoe to drop.
After three hours on the different powersupply one of the drives (with
no SATA connection) has spontaneously started clunking every few seconds
and has twice spun down and up again. When I picked it up to read the
label it stopped clunking briefly then resumed.
Samsung SP2504C 250GB
Definitively FC5 is nothing to do with my HDD unhappiness.
-Andy
8:07 a.m.
New subject: Crappy HDDs Killing Computers
On Tue, 2006-08-22 at 13:04 +0100, Andy Green wrote:
Chong Yu Meng wrote:
> Let's all wait for the other shoe to drop.
After three hours on the different powersupply one of the drives (with
no SATA connection) has spontaneously started clunking every few seconds
and has twice spun down and up again. When I picked it up to read the
label it stopped clunking briefly then resumed.
Samsung SP2504C 250GB
Samsung ???? Auuugh!! There are four drive manufacturers I try to stay
away from: IBM/Hitachi, Maxtor, Fujitsu and Samsung. I guess that just
leaves Seagate and Western Digital. Incidentally, those are the 2 oldest
drives that I still have that still work. A 80MB Western Digital
(remember when 80MB was huge ?) and a 18GB Seagate SCSI in a external
casing.
Of course, my experience is completely subjective and a lot depends on
your luck. From my previous experience as tech support for a hardware
reseller (RMA) many years ago, I learned that there are batches
occasionally that slip through Quality Control either because of a new
undetected fault or a change in their Quality testing. If you happen to
be one of the "victims", it can be especially frustrating because other
users may testify to the reliability of the brand, when you are tearing
your hair out wondering if you are crazy.
Also, there are the hard disks that "fall off the back of the truck" --
hijacked shipments that may or may not contain good drives. It's
probably a smaller problem in the West, but it used to be a big problem
in Asia.
--
Pascal Chong
email: chongym(a)cymulacrum.net
web: http://cymulacrum.net
pgp: http://cymulacrum.net/pgp/cymulacrum.asc
"La science ne connaît pas de frontière parce que la connaissance
appartient à l’humanité. et que c’est la flamme qui illumine le monde."
-- Louis Pasteur
8:18 a.m.
New subject: Crappy HDDs Killing Computers
Chong Yu Meng wrote:
Samsung ???? Auuugh!! There are four drive manufacturers I try to
stay
away from: IBM/Hitachi, Maxtor, Fujitsu and Samsung. I guess that just
Sigh, Samsung HDD in this Samsung laptop to look forward to getting to
know better then. And FWIW the drive that fell over and made be buy the
Samsung in a hurry was a Western Digital Raptor WD740 74GB 10KRPM job.
But that lasted the best part of a year in the box.
leaves Seagate and Western Digital. Incidentally, those are the 2
oldest
drives that I still have that still work. A 80MB Western Digital
(remember when 80MB was huge ?) and a 18GB Seagate SCSI in a external
casing.
Heh seems not too long ago (20 years +) I remember buying a surplus 10MB
HDD, huge, heavy beautifully engineered thing with an unfathomable
proprietary interface, wonderful coloured twisted-pair ribbon cables.
Errr.. must've dozed off there for a minute.
Also, there are the hard disks that "fall off the back of the
truck" --
hijacked shipments that may or may not contain good drives. It's
No I doubt it fell off a truck, although in fact it is behaving rather
like it did. I got it in a hurry from a "PC World" store here in the
UK. Anyway I used to work out of a laptop for years and am remembering
all the good points of that at the moment so I guess the desktop was a
failed experiment. I guess you must be right and I was in the wrong
store at the wrong time buying the wrong product.
-Andy
11:04 a.m.
New subject: Crappy HDDs Killing Computers
On Tue, 2006-08-22 at 14:18 +0100, Andy Green wrote:
I guess you must be right and I was in the wrong
store at the wrong time buying the wrong product.
In Singapore, a lot of us believe that such a confluence of bad luck
means that good luck must be around the corner! Buy yourself a lottery
ticket ! Or give me the last four digits of the hard disk's serial
number and I will buy myself a lottery ticket ! ;)
--
Pascal Chong
email: chongym(a)cymulacrum.net
web: http://cymulacrum.net
pgp: http://cymulacrum.net/pgp/cymulacrum.asc
"La science ne connaît pas de frontière parce que la connaissance
appartient à l’humanité. et que c’est la flamme qui illumine le monde."
-- Louis Pasteur
11:48 a.m.
New subject: Crappy HDDs Killing Computers
Uh, Seagate just bought Maxtor, fyi.
I have had a lot of trouble with shuttle machines and seagate drives that
get hot and BURN UP due to bad design of the case which doesn't dissipate
heat. It only happens with the second drive installed. I have replaced
some seagates with Maxtor drives and I have not had the problem. I am
convinced that it is not a 'bad batch' of drives since the machines were
bought over a time period of about a year.
That's my hard drive story of the day.
Marc
On 8/22/06, Chong Yu Meng <chongym(a)cymulacrum.net> wrote:
On Tue, 2006-08-22 at 14:18 +0100, Andy Green wrote:
> I guess you must be right and I was in the wrong
> store at the wrong time buying the wrong product.
>
In Singapore, a lot of us believe that such a confluence of bad luck
means that good luck must be around the corner! Buy yourself a lottery
ticket ! Or give me the last four digits of the hard disk's serial
number and I will buy myself a lottery ticket ! ;)
--
Pascal Chong
email: chongym(a)cymulacrum.net
web: http://cymulacrum.net
pgp: http://cymulacrum.net/pgp/cymulacrum.asc
"La science ne connaît pas de frontière parce que la connaissance
appartient à l'humanité. et que c'est la flamme qui illumine le monde."
-- Louis Pasteur
--
fedora-list mailing list
fedora-list(a)redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
11:50 a.m.
New subject: Crappy HDDs Killing Computers
On Tue, 22 Aug 2006, Marc wrote:
Uh, Seagate just bought Maxtor, fyi.
I have had a lot of trouble with shuttle machines and seagate drives that
get hot and BURN UP due to bad design of the case which doesn't dissipate
heat. It only happens with the second drive installed. I have replaced
some seagates with Maxtor drives and I have not had the problem. I am
convinced that it is not a 'bad batch' of drives since the machines were
bought over a time period of about a year.
That's my hard drive story of the day.
I used to heat my apartment with a 4.2gig Seagate SCSI drive.
No, I am not kidding. It ran that hot. (And it was a small apartment.)
--
"Oh, Joel Miller, you've just found the marble in the oatmeal. You're a
lucky, lucky, lucky little boy. 'Cause you know why? You get to drink
from... the FIRE HOOOOOSE!"
- The Stanley Spudoski guide to mailing list administration
12:38 p.m.
New subject: Crappy HDDs Killing Computers
alan wrote:
I used to heat my apartment with a 4.2gig Seagate SCSI drive.
No, I am not kidding. It ran that hot. (And it was a small apartment.)
This was a 5-1/4" full height drive, right? I used to use one in an
external case as a foot warmer. (I think I still have it in storage.)
Mikkel
--
Do not meddle in the affairs of dragons,
for thou art crunchy and taste good with Ketchup!
12:58 p.m.
New subject: Crappy HDDs Killing Computers
On Tue, 2006-08-22 at 12:38 -0500, Mikkel L. Ellertson wrote:
alan wrote:
>
> I used to heat my apartment with a 4.2gig Seagate SCSI drive.
To be honest i really think that Seagate is much much better than
Maxtor, i think Maxtor is crap, maybe because i have had some really
rotten luck with Maxtor, from bad sectors to total failure while Seagate
has been the most reliable for me.
>
> No, I am not kidding. It ran that hot. (And it was a small apartment.)
>
This was a 5-1/4" full height drive, right? I used to use one in an
external case as a foot warmer. (I think I still have it in storage.)
Mikkel
--
Do not meddle in the affairs of dragons,
for thou art crunchy and taste good with Ketchup!
1:22 p.m.
New subject: Crappy HDDs Killing Computers
Roger wrote:
On Tue, 2006-08-22 at 12:38 -0500, Mikkel L. Ellertson wrote:
> alan wrote:
>> I used to heat my apartment with a 4.2gig Seagate SCSI drive.
To be honest i really think that Seagate is much much better than
Maxtor, i think Maxtor is crap, maybe because i have had some really
rotten luck with Maxtor, from bad sectors to total failure while Seagate
has been the most reliable for me.
>> No, I am not kidding. It ran that hot. (And it was a small apartment.)
>>
> This was a 5-1/4" full height drive, right? I used to use one in an
> external case as a foot warmer. (I think I still have it in storage.)
>
> Mikkel
> --
>
> Do not meddle in the affairs of dragons,
> for thou art crunchy and taste good with Ketchup!
>
Just a little note, Seagate now owns Maxtor.
John
6:49 p.m.
New subject: Crappy HDDs Killing Computers
From: "Roger" <roger(a)wack.co.za>
On Tue, 2006-08-22 at 12:38 -0500, Mikkel L. Ellertson wrote:
> alan wrote:
> >
> > I used to heat my apartment with a 4.2gig Seagate SCSI drive.
To be honest i really think that Seagate is much much better than
Maxtor, i think Maxtor is crap, maybe because i have had some really
rotten luck with Maxtor, from bad sectors to total failure while Seagate
has been the most reliable for me.
LONG ago I learned the value of backups, particularly live drive
backups. All information that is critical is in multiple places.
Then I decided that was not enough. You might say that these days
I am a RAID evangelist. So far I have not lost anything detectable
to a RAID array. I sweat blood until the RAID is whole again. But
both my partner and I are at zero loss lately thanks to RAID arrays.
(Of course, don't discount the RAID board itself as a potential
error source. Once I nailed it to the Promise card they were VERY
nice about replacing it. But I am ashamed of how long it took me
to figure out that the error, at the same block on 4 drives in
succession, was due to that card and not the drives. I was starting
to get annoyed with the Maxtor failures when they were all in
reality the RAID card. Go figure. Sometimes I am remarkably slow
to figure that sort of thing out. <sigh>)
{^_^}
6:42 p.m.
New subject: Crappy HDDs Killing Computers
And these days I rather like both of them. I've not lost data to a
Maxtor drive in ages. My partner used Seagate and has done pretty
well in that regard, too. He's had more failures than I have. But
RAID works. (And I think his failures were power supply. Mine were
solely related to a defective Promise RAID card that installed errors
in identical places on two disks in the RAID. Maxtor replaced them
no questions. And once Promise replaced the card it's been perfect
ever since.) We have a PILE of old Western Digital failures if anyone
is interested in 2G or 3G drives. Ever since we've eschewed them. We
also have a similar pile of IBM DeathStars. And I do have a pair of
80G Samsungs that are quite happily working as the drives in the
Linux firewall machine for quite some years now 24/7/365.24 less
maybe eight hours of hardware reconfiguration time on the machine
a year ago.
These days I look to the guarantee and purchase the premium drives
from Seagate or Maxtor. Those seem to be built to last forever. The
commodity drives die like flies no matter who makes them.
{^_^}
----- Original Message -----
From: "Marc " <linuxr(a)gmail.com>
Uh, Seagate just bought Maxtor, fyi.
I have had a lot of trouble with shuttle machines and seagate drives that
get hot and BURN UP due to bad design of the case which doesn't dissipate
heat. It only happens with the second drive installed. I have replaced
some seagates with Maxtor drives and I have not had the problem. I am
convinced that it is not a 'bad batch' of drives since the machines were
bought over a time period of about a year.
That's my hard drive story of the day.
Marc
On 8/22/06, Chong Yu Meng <chongym(a)cymulacrum.net> wrote:
On Tue, 2006-08-22 at 14:18 +0100, Andy Green wrote:
> I guess you must be right and I was in the wrong
> store at the wrong time buying the wrong product.
>
In Singapore, a lot of us believe that such a confluence of bad luck
means that good luck must be around the corner! Buy yourself a lottery
ticket ! Or give me the last four digits of the hard disk's serial
number and I will buy myself a lottery ticket ! ;)
--
Pascal Chong
email: chongym(a)cymulacrum.net
web: http://cymulacrum.net
pgp: http://cymulacrum.net/pgp/cymulacrum.asc
"La science ne connaît pas de frontière parce que la connaissance
appartient à l'humanité. et que c'est la flamme qui illumine le monde."
-- Louis Pasteur
--
fedora-list mailing list
fedora-list(a)redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
--------------------------------------------------------------------------------
--
fedora-list mailing list
fedora-list(a)redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
4:23 p.m.
New subject: Crappy HDDs Killing Computers
On Tue, 2006-08-22 at 21:07 +0800, Chong Yu Meng wrote:
Samsung ???? Auuugh!! There are four drive manufacturers I try to
stay
away from: IBM/Hitachi, Maxtor, Fujitsu and Samsung.
Actually Maxtor drives work great - you just need to know the
algorithm:
1. Buy drive at local office/computer supply store
2. Have it fail 2 weeks later
3. RMA it to Maxtor
4. Get replacement that works forever.
This has been my consistent experience with Maxtor.
You just need to be prepared to take several
weeks to actually get a working disk :-).
8:10 p.m.
New subject: Crappy HDDs Killing Computers
On Wed, 2006-08-23 at 17:23 -0400, Tom Horsley wrote:
Actually Maxtor drives work great - you just need to know the
algorithm:
1. Buy drive at local office/computer supply store
2. Have it fail 2 weeks later
3. RMA it to Maxtor
4. Get replacement that works forever.
This has been my consistent experience with Maxtor.
You just need to be prepared to take several
weeks to actually get a working disk :-).
Hmm, that seems to imply a couple of things:
1. Bad handling of the drives during transport.
2. Bad handling of the drives at the store.
Some people do have no idea about being gentle with them, nor taking
anti-static precautions.
--
(Currently running FC4, occasionally trying FC5.)
Don't send private replies to my address, the mailbox is ignored.
I read messages from the public lists.
11:45 a.m.
New subject: Crappy HDDs Killing Computers
On Tue, 2006-08-22 at 13:04 +0100, Andy Green wrote:
After three hours on the different powersupply one of the drives
(with
no SATA connection) has spontaneously started clunking every few
seconds and has twice spun down and up again. When I picked it up to
read the label it stopped clunking briefly then resumed.
Just for interest's sake:
Was it mounted somewhere with decent air flow for heat dissipation?
Was it bolted to a metal chassis that would help with heat dissipation?
Was it grounded?
Were anti-static precautions taken during handling?
--
(Currently running FC4, occasionally trying FC5.)
Don't send private replies to my address, the mailbox is ignored.
I read messages from the public lists.
1:03 p.m.
New subject: Crappy HDDs Killing Computers
Tim wrote:
On Tue, 2006-08-22 at 13:04 +0100, Andy Green wrote:
> After three hours on the different powersupply one of the drives (with
> no SATA connection) has spontaneously started clunking every few
> seconds and has twice spun down and up again. When I picked it up to
> read the label it stopped clunking briefly then resumed.
Just for interest's sake:
Was it mounted somewhere with decent air flow for heat dissipation?
No, it was resting on top of a plastic box in a room during this test.
But on picking up the HDD it only felt quite warm, I could hold it easily.
While in the case there were multiple fans and it was in the appropriate
slot. Further the second HDD to fail did not last out a week in the
same position that the first lasted most of year.
Was it bolted to a metal chassis that would help with heat
dissipation?
Was it grounded?
Yeah it was via a Class 1 PSU where 0V == Earth Ground.
Were anti-static precautions taken during handling?
No, I have never had a failure of anything I could put down to static,
and that includes unmounted chips much more sensitive than a HDD IO
connector. Certainly when it was in the case there is zero chance of
static causing the trouble since the case is earthed metal and the HDD
does not come to the outside world.
Recently I took a design of mine through testing for compliance to
various standards, and it survived 16kV air discharges and one directly
to the case by accident, without special precautions. In a climate like
England, static is not a real consideration with modern ICs.
-Andy
8:22 p.m.
New subject: Crappy HDDs Killing Computers
Tim:
> Just for interest's sake:
>
> Was it mounted somewhere with decent air flow for heat dissipation?
Andy Green:
No, it was resting on top of a plastic box in a room during this
test.
But on picking up the HDD it only felt quite warm, I could hold it
easily.
Yes, I'd have said that's fine. I'd say they've got to get very hot, or
be overcooled, before you get concerned about temperature. Any drive
that runs cool or luke warm ought to be fine (else it's never going to
survive in the real world).
> Were anti-static precautions taken during handling?
No, I have never had a failure of anything I could put down to
static,
and that includes unmounted chips much more sensitive than a HDD IO
connector.
This is a hard one to really quantify. Much static-caused damage is not
instant death to the device. Later on you just do not equate what was
done before as the cause of something dying at a different time. And
while it is far less likely that a complete device, rather than a loose
IC, might be damaged by static, it is still a possibility. It can also
be that static damaged it long before you ever got your hot little hands
on it (manufacturing, transport, shops, etc.).
Other things that kill drives can be: Vibrations, they should be bolted
firmly in place, and nothing else should vibrate the chassis. Tortional
stress - you don't want a chassis that twists or squashes a drive.
Certainly when it was in the case there is zero chance of static
causing the trouble since the case is earthed metal and the HDD
does not come to the outside world.
Yes, I'd say that's a near certainty. However, there have been cases of
people static-zapping computers (the working, installed, computer; not
individual devices out of the box), simply by working around them.
Recently I took a design of mine through testing for compliance to
various standards, and it survived 16kV air discharges and one
directly to the case by accident, without special precautions. In a
climate like England, static is not a real consideration with modern
ICs.
Hmm, I don't know I'd be so sure about the climate protecting you.
Sure, I don't expect a built computer sitting in a rack to suffer
problems, but I'd not walk around with a CMOS chip in my fingers
expecting that it'd survive better just because it was in England. They
just don't have that stiff upper lip. ;-)
--
(Currently running FC4, occasionally trying FC5.)
Don't send private replies to my address, the mailbox is ignored.
I read messages from the public lists.
11:57 p.m.
New subject: Crappy HDDs Killing Computers
From: "Tim" <ignored_mailbox(a)yahoo.com.au>
Other things that kill drives can be: Vibrations, they should be
bolted
firmly in place, and nothing else should vibrate the chassis. Tortional
stress - you don't want a chassis that twists or squashes a drive.
If you are in a high vibration environment the drives really should be
mounted in shock mounts if you have the option. Solidly mounting them
to the chassis simply conveys the shocks directly to the delicate
insides of the drives.
{^_^}
2:36 a.m.
New subject: Crappy HDDs Killing Computers
Tim:
> Other things that kill drives can be: Vibrations, they should
be bolted
> firmly in place, and nothing else should vibrate the chassis. Tortional
> stress - you don't want a chassis that twists or squashes a drive.
jdow:
If you are in a high vibration environment the drives really should
be
mounted in shock mounts if you have the option. Solidly mounting them
to the chassis simply conveys the shocks directly to the delicate
insides of the drives.
Oh definitely. Though sometimes you put a drive into a situation where
you never expected vibration to be an issue:
I've had a fan go wonky in a box last week that had the whole chassis
thrumming along with it, to the point where the drive started
misbehaving. Luckily, in this case, all it did was show up a fault in a
poor hard drive power plug - one of those crappy 4-pin molex plastic
connectors - a loose gripping contact. But when you consider the tiny
dimensions of the head skimming above the platter, continuous (and
otherwise) small vibrations in the chassis can be a significant thing.
--
(Currently running FC4, in case that's important to the thread)
Don't send private replies to my address, the mailbox is ignored.
I read messages from the public lists.
11:21 a.m.
Tim wrote:
Robin Laing:
>>Depending on the hardware, running Linux on it can push cheap hardware
>>past it's limits.
>>
>>Multitasking/mutithreading can load a system more than just the old
>>windows way of doing things. This causes the processors and power
>>supplies to work harder.
I've read this before, but I find it hard to believe. I regularly see
Windows systems where the CPU maxes out a lot, doing ordinary tasks.
I've seen many where the resources are used almost to the full almost
all of the time. But on the same hardware using Linux, instead, I've
usually found that it spends most of its time being idle. Sure, you can
have some heavy intensive tasks, but that doesn't tend be the normal way
that Linux runs.
>>FC5 won't be burning the computers out but it could be showing the
>>weaknesses of the hardware.
As could any strenuous task.
I have seen problems occur on machines around work here that were moved
to Linux due to load issues. On the Linux boxes we normally run BOINC
as part of our research. Of course we put the machines under higher load.
Of course, most of the statements about multitasking, etc are based on
older versions of Windows and slower processors.
I just remembered that there was a problem with some motherboards and
bad components a few years back. I think it was a bad batch of
capacitors. It was also across manufacturers. Check Google on this.
BTW, we use mainly Fedora on all our Linux machines. Most machines are
FC4, some are FC5. The odd FC3 machine still runs. A couple of RedHat
machines for some specialized commercial software.
--
Robin Laing
11:44 a.m.
On Tue, 2006-08-22 at 10:21 -0600, Robin Laing wrote:
Of course, most of the statements about multitasking, etc are based
on
older versions of Windows and slower processors.
My experience has been that Windows XP, arguably the most stable Windows
version ever (haven't tried Vista) seems to stress my hardware more than
older versions of Windows (Windows 2000 Professional) running on older
hardware. I don't recall my hard disks thrashing as much on Windows
2000. YMMV
--
Pascal Chong
email: chongym(a)cymulacrum.net
web: http://cymulacrum.net
pgp: http://cymulacrum.net/pgp/cymulacrum.asc
"La science ne connaît pas de frontière parce que la connaissance
appartient à l’humanité. et que c’est la flamme qui illumine le monde."
-- Louis Pasteur
1:57 a.m.
Replies Interwoven:
On 8/18/06, Robert Gann <gannr(a)hughes.net> wrote:
Hi Robert Gann!! Glad to see you here!!
> I have a most distressing problem. I am a faculty member in the computer
> and information sciences department at a liberal arts college.
>
It was easy for me to find that. This may be more of your problems
than you may first imagine (but then some call me paranoid).
> We have a lab with a number of computers with two AMD 32 bit CPUs and a few
> computers with two AMD 64 bit CPUs.
>
> Two years ago we ran FC4 on both types of these computers with no problem.
> However, attempting to install FC5 has killed several computers. We have
> now burned out two of the 64 bit computers and two of the 32 bit computers.
> Generally, the behavior is that they go a ways into the installation process
> and hang. After retrying the installation another time, the computer is
> basically dead; when I tried a Knoppix CD-ROM one of them could not get past
> the screen telling you to press a key to boot Knoppix. One of the 64 bit
> computers went "poof", and I smelled smoke. It is,of course, completely
> dead.
STEP ONE: CONFIRM THE SOURCE OF YOUR ISOs!!!!! Do the checksums
match. Can you trust the sources of your checksums. Do you have pure
FC5 or do you have something designed to hurt you???!?!?
If you used source code, same story different ways to confirm purity!
STEP TWO: CONFIRM THE INTEGRITY OF THE BIOS!!!! I mean find out how
to copy it from the flash memory and do a compair with what should be
there!! I can think of a couple of things I could do at a Kernel or
driver level to smoke a computer (busy it with pci, agp, epci issues
or some such) but at a bios level I can turn up the voltage,
overclock, and then loop it!!! One of Paul E. Proctor's points about
Computer Security (The Practical Intrusion Detection Handbook p.100
1999 CSI/FBI CCSS) is that only 20% of computer abuse comes from the
outside, the rest is from within.
STEP THREE: Test the CMOS batteries (obvious reasons). Make sure the
holder is clean.
Step Four: Chart out as well as you can remember what you did. You
need to be objective and accurate ("CSI" mode if you will). Order of
steps, software loaded and run / running, power, net, phone, TV,
Video, all things connected!! Were the boxes removed from their
normal place? Was a different source for net, disk, or other "hose"
connected at the time? Was the cover on? Was a different source of
power used by the computer or ANYTHING connected to the computer.
Step five: Check the local environment. Power filters ok?
Temperature, humidity, source of dripping water nearby, smoke,
chemicals, physical shock. Any other computers in the lab down?
When? Do the computers "glitch" once in a while, Do they do so at a
predictable time? If you are rural, or if you are city near
construction you can have some Hugh power spikes reaching you. Is
there a source of RFI (Broadcast station, HAM, Linear CB, Leakey
Microwave, Linear Wireless ) or other EMI (Big Printer, Refrigeration,
other electric motor, dimmer switch, etc...)?
Step six: If you have not found any other solid suspect for the
smoking of your computers, considering that your organization and
person have benefited from this OSS please do consider it your
responsibility to establish communication with the FC5 Kernel and
Driver development teams. They need to know and we all need them to
know.
>
> I've used Google to see if I could find out about a problem to no avail.
> We're getting pretty desperate. Can anyone offer any guidance? At this
> point, we're even afraid to try another distribution. We're starting to
> install Windows to get the remaining computers working.
>
> I'll be glad to provide more details if they will help.
>
> Thank you in advance..
>
> --
> fedora-list mailing list
> fedora-list(a)redhat.com
> To unsubscribe:
> https://www.redhat.com/mailman/listinfo/fedora-list
Robert, I regret to hear about the losses you all have seen. I hope
this is "guidance" of some sort. It is what I would be doing.
May you find the source!!!
Tod
6485
days inactive
6497
days old
41 comments
23 participants
participants (23)
-
Alan -
Andy Green -
Benjamin Franz -
Chong Yu Meng -
Chris Mohler -
Craig White -
Ed Greshko -
Gilboa Davara -
Hex Star -
Jack Gates -
jdow -
John Wendel -
Marc M -
Mikkel -
Myles Green -
Ric Moore -
Robert Gann -
Robert L Cochran -
Robin Laing -
Roger -
Tim -
Tod Merley -
Tom Horsley