http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=893808

Can you tell me what exit code 5 is?
<a href="http://tinyurl.com/9hemz"><img src="http://www.boincsynergy.com/images/stats/comb-1441.jpg"></img></a>

-5 is a general purpose code that covers various problems.

Les

so it could be absolutely anything?
<a href="http://tinyurl.com/9hemz"><img src="http://www.boincsynergy.com/images/stats/comb-1441.jpg"></img></a>

Not everything. There are other error codes.
One problem covered by -5 is when a negative pressure is calculated in one of the many cells of the model.
This may be caused by a hardware problem, from a dust covered heatsink on the cpu chip, to a power supply that's not quite up to it. And several others in between.

Have you tried the hardware/software tests from the lists in the community forum?

Something that you could try if you still have the zip files from the model, or the raw files if it crashed early.

In the file yabsd.out there is, apparently, a message about the cause of the crash.
I haven't worked out how to unzip in XP, or I'd look at my two that crashed last year. I've got a program that will unzip anything, but it's DOS, from The Good Old Days of Win 3.11.

Les

&gt; Not everything. There are other error codes.
&gt; One problem covered by -5 is when a negative pressure is calculated in one of
&gt; the many cells of the model.
&gt; This may be caused by a hardware problem, from a dust covered heatsink on the
&gt; cpu chip, to a power supply that's not quite up to it. And several others in
&gt; between.
&gt;
&gt; Have you tried the hardware/software tests from the lists in the community
&gt; forum?
&gt;
&gt;
&gt;
&gt;
&gt;

no...can you point me to those?

Negative pressure = people explode?
<a href="http://tinyurl.com/9hemz"><img src="http://www.boincsynergy.com/images/stats/comb-1441.jpg"></img></a>

<a href="http://www.climateprediction.net/board/viewtopic.php?t=2126">Hardware specific test</a>
<a href="http://www.climateprediction.net/board/viewtopic.php?t=2124">Basic hardware maintenance</a>

I spent too much time looking for posts on negative pressure by Carl, so I deleted my text.

Les

Check the Wiki

<a href="http://boinc-doc.net/boinc-wiki/index.php?title=Unrecoverable_error_for_result_%27%28result%29%27_%28_-_exit_code_-5_%280xffffffb%29%29">WIKI</a>

<a href="http://boinc-doc.net/boinc-wiki/index.php?title=Main_Page">Link to Unoffical Wiki for <b>BOINC</b></a>

&gt; Check the Wiki
&gt;

"These errors may be the result of a hardware error, but are more likely because of a problem with the Science Application software and the Work Unit that it is trying to process."

Quote from the Wiki above. I would think in this case in this post, given the large number of errors this host has had of this type, that a problem with the application software and WU could be ruled out.

&gt; Quote from the Wiki above. I would think in this case in this post, given the
&gt; large number of errors this host has had of this type, that a problem with the
&gt; application software and WU could be ruled out.

No.

I have not completed a model on my PowerMac in like forever. I have no idea what the cause might me, but all other eligible projects have no problem. It could be an incompatibility of "Tiger" with CPDN. Or it could just be a bad compile.

The single most likely cause, in Paul's biased opinion, is hardware instability because of overclocking or just plain "cheap" hardware.

Unfortunately, isolating a problem like this is very difficult because of the incredible complexity of the systems we use.

I find it hard to believe that any of my hardware could be faulty when none of the hardware in my system is older that 1 yr. old (and the time they've been in there no errors so I doubt any manufacturing error). Plus, I am running 3 other BOINC projects all fine. And I'm using BOINC 4.43/4.44/4.45 so it may be the new version of boinc.

CPU: About 6 months old
Motherboard: About 6 months old
Power Source (Dunno what it's called) : About 9 months old
Graphics Card: ABout 1 year old
RAM: About 6 months old

Also, 4 WU's were abandoned (sent but never received) 3 show computing error two of which had error code -1. So, this is the first time I've had a -5.
<a href="http://tinyurl.com/9hemz"><img src="http://www.boincsynergy.com/images/stats/comb-1441.jpg"></img></a>

The thing about hardware is if it's going to fail, it usually fails quickly. If it doesn't fail quickly, then it's going to usually last a long long time.

Do some generic hardware tests. Memory, processor, etc. It's never a bad idea to do this, and it may find something you didn't know. Many times hardware is slightly bad, but because we do not use the full potential of the hardware, it goes unnoticed. When you use a program like Boinc that does use a lot of processing power, then you find these errors.

Do it while it's under warrenty!! (-:


<img src="http://seti.mundayweb.com/stats.php?userID=344&amp;trans=off">
<img src="http://boinc.mundayweb.com/one/stats.php?userID=584&amp;trans=off">

&gt; The thing about hardware is if it's going to fail, it usually fails quickly.
&gt; If it doesn't fail quickly, then it's going to usually last a long long time.
&gt;
&gt; Do some generic hardware tests. Memory, processor, etc. It's never a bad
&gt; idea to do this, and it may find something you didn't know. Many times
&gt; hardware is slightly bad, but because we do not use the full potential of the
&gt; hardware, it goes unnoticed. When you use a program like Boinc that does use
&gt; a lot of processing power, then you find these errors.
&gt;
&gt; Do it while it's under warrenty!! (-:
&gt;


but that's my point is that I've only had one WU exit under -5 and all my other projects work fine. I'll do the tests a little later just in case though.
<a href="http://tinyurl.com/9hemz"><img src="http://www.boincsynergy.com/images/stats/comb-1441.jpg"></img></a>

&gt; but that's my point is that I've only had one WU exit under -5 and all my
&gt; other projects work fine. I'll do the tests a little later just in case
&gt; though.

But with the size of CPDN, it may be the item that puts you over the edge.


<img src="http://seti.mundayweb.com/stats.php?userID=344&amp;trans=off">
<img src="http://boinc.mundayweb.com/one/stats.php?userID=584&amp;trans=off">

Hi Ilyanep

I just wondered whether your machine could possibly be an Athlon...........and yes it is.

Do run the stability tests etc first but if, as has happened to several other would-be crunchers, your machine performs to Olympic standard, we may be back to the old question of Athlons' number-crunching technique being less compatible with the boinc cpdn software than Pentiums.

Don't ask me the mathematical details, but this has been discussed ad infinitum before - something to do with floating points. This can lead to calculation errors which get the -5 code and mean that negative pressure has been generated. Maybe we would all implode rather than explode. Anyway, the weather has gone wrong, so it's just as well for the researchers that these models abort.

I eventually allowed my Athlon to concede defeat in the battle with boinc cpdn, and reverted to crunching classic cpdn, where a few of us will remain until cpdn boinc becomes kinder to Athlons. (Classic works brilliantly and if it's good enough for the Open University course, it's good enough for me.)

Obviously, not all Athlons encounter this problem, and some members have crunched hundreds of models with them. But -5 errors seem to be generated more frequently on Athlons than Pentiums.

Now if someone can tell me that this problem has been resolved, I'll try cpdn boinc again.
__________________________________________________

&gt; Hi Ilyanep
&gt;
&gt; I just wondered whether your machine could possibly be an Athlon...........and
&gt; yes it is.
&gt;
&gt; Do run the stability tests etc first but if, as has happened to several other
&gt; would-be crunchers, your machine performs to Olympic standard, we may be back
&gt; to the old question of Athlons' number-crunching technique being less
&gt; compatible with the boinc cpdn software than Pentiums.
&gt;
&gt; Don't ask me the mathematical details, but this has been discussed ad
&gt; infinitum before - something to do with floating points. This can lead to
&gt; calculation errors which get the -5 code and mean that negative pressure has
&gt; been generated. Maybe we would all implode rather than explode. Anyway, the
&gt; weather has gone wrong, so it's just as well for the researchers that these
&gt; models abort.
&gt;
&gt; I eventually allowed my Athlon to concede defeat in the battle with boinc
&gt; cpdn, and reverted to crunching classic cpdn, where a few of us will remain
&gt; until cpdn boinc becomes kinder to Athlons. (Classic works brilliantly and if
&gt; it's good enough for the Open University course, it's good enough for me.)
&gt;
&gt; Obviously, not all Athlons encounter this problem, and some members have
&gt; crunched hundreds of models with them. But -5 errors seem to be generated more
&gt; frequently on Athlons than Pentiums.
&gt;
&gt; Now if someone can tell me that this problem has been resolved, I'll try cpdn
&gt; boinc again.
&gt; __________________________________________________
&gt;
&gt;
&gt;

this may or may not be my problem. That's quite fascinating. I'd rather not go to classic CPDN though because I have 3 other BOINC projects running.
<a href="http://tinyurl.com/9hemz"><img src="http://www.boincsynergy.com/images/stats/comb-1441.jpg"></img></a>

&gt; Obviously, not all Athlons encounter this problem, and some members have
&gt; crunched hundreds of models with them. But -5 errors seem to be generated more
&gt; frequently on Athlons than Pentiums.

My AMD64 FX-53 hasn't encountered any problems. (Currently on hiatus from production CPDN while running alpha)

As we speak I am 21 hours into a CPDN WU no problem. I'm starting to think this may have been a bug withing 4.43 and 4.44?
<a href="http://tinyurl.com/9hemz"><img src="http://www.boincsynergy.com/images/stats/comb-1441.jpg"></img></a>