http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3300476



One guy takes 15 seconds... and its not invalid? How is this possible??

> http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3300476
>
>
>
> One guy takes 15 seconds... and its not invalid? How is this possible??

Imagine a WU with practically no signals ie. only background noise far below normal levels of signals so no really elaborate calculations (FFT) have to be done...
Even no result can scientifically be valuable.

Greetings ;-))

---

> > http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3300476
> >
> >
> >
> > One guy takes 15 seconds... and its not invalid? How is this possible??
>
> Imagine a WU with practically no signals ie. only background noise far below
> normal levels of signals so no really elaborate calculations (FFT) have to be
> done...
> Even no result can scientifically be valuable.
>
> Greetings ;-))
>
>
>


But arent 3 identical WU's sent out to 3 people?

---

> > http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3300476
> >
> >
> >
> > One guy takes 15 seconds... and its not invalid? How is this possible??
>
> Imagine a WU with practically no signals ie. only background noise far below
> normal levels of signals so no really elaborate calculations (FFT) have to be
> done...
> Even no result can scientifically be valuable.

Except that isn't what's being discussed. The "problem" is that three similar systems returned results in (15), (18,000), and (20,000) seconds. One of those values does not seem reasonable for a system having had returned a correct result.

trane

---

<p>
[url=http://setiweb.ssl.berkeley.edu/workunit.php?wuid=2974310 ]
2974310 [/url]
</p>


If you think thats odd check out this one I was recently credited with
I'm the host 321394
this was a work unit that failed to upload

---

visit boinc@Australia

Gotta like it when you get credit for doing nothing. LOL

> <p>
> [url=http://setiweb.ssl.berkeley.edu/workunit.php?wuid=2974310 ]
> 2974310 [/url]
> </p>
>
>
> If you think thats odd check out this one I was recently credited with
> I'm the host 321394
> this was a work unit that failed to upload
>

---

I got a few of those as well.

Example :
<p>
3793818
</p>

My Result ( http://setiweb.ssl.berkeley.edu/result.php?resultid=15247355 ) Claimed 0 Credits due to 0 Seconds runtime, with an Error Message :

"SETI@Home Informational message -9 result_overflow
NOTE: The number of results detected exceeds the storage space allocated."

Since the other CPU's took more than 100 Seconds to find that out (?), they naturally claimed higher Credit.
Since the Results are identical in the end, I got Credit for "nothing" as well ;)

---

>
> My Result ( http://setiweb.ssl.berkeley.edu/result.php?resultid=15247355 )
> Claimed 0 Credits due to 0 Seconds runtime, with an Error Message :
>
> "SETI@Home Informational message -9 result_overflow
> NOTE: The number of results detected exceeds the storage space allocated."
>

This isn't really an error, and isn't counted as such, so the validator is validating these wu just like any other wu. It's just like in "classic" a wu with too much noise so terminates very fast.

As for only reporting 0 seconds, this is wrong. It looks like BOINC doesn't correctly coiunt the 1st minute of a wu or something, therefore some fast wu will report as 0 seconds. Depending on the 2 other crunchers you're sometimes getting credit for these, and sometimes not. In any case it's not a big problem since gives very little credit, and it's not so many of these wu.

> <p>
> [url=http://setiweb.ssl.berkeley.edu/workunit.php?wuid=2974310 ]
> 2974310 [/url]
> </p>
>
>
> If you think thats odd check out this one I was recently credited with
> I'm the host 321394
> this was a work unit that failed to upload
>

haha yea, thats defenately an error in the validator.

---

Here's one that got me a bit frustrated: I got sent a WU that was already crunched by two CPUs. THe second CPU claimed something like 40 cr., while the first claimed 0. When I finished it, I claimed 30, but everyone got 0. Now I can't even find the WU in the list! [grrr!] The WU# is 2579151.

Why would SETI send out a WU that has a 0 cr. claim - Isn't that a sign that there's nothing in the WU to look for?

> Here's one that got me a bit frustrated: I got sent a WU that was already
> crunched by two CPUs. THe second CPU claimed something like 40 cr., while the
> first claimed 0. When I finished it, I claimed 30, but everyone got 0. Now I
> can't even find the WU in the list! [grrr!] The WU# is 2579151.
>
> Why would SETI send out a WU that has a 0 cr. claim - Isn't that a sign that
> there's nothing in the WU to look for?
>

Many results is wrongly marked as "success" even example was download-error, but the validator will correctly mark these invalid. Also, it looks like BOINC doesn't correctly count the 1st minute or something, so if a wu terminated with -9 "too many signals" it can be valid even with zero reported time.

But "everyone got 0" indicates either 2 claimed zero, or there was some rror-results alongside the "success", meaning very likely too many errors so wu never tried validated at all.

can anyone tell me what happened here? http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3819102

2 others get credit, i got 0.00 It's invalid? i'm wondering if this is because maybe windows blue screened at the wrong moment?

It was an invalid return for your returned result, why it was invalid I don't know ...

> can anyone tell me what happened here?
> http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3819102
>
> 2 others get credit, i got 0.00 It's invalid? i'm wondering if this is because
> maybe windows blue screened at the wrong moment?
>
Mighty Mik

I would suspect that the Blue screen would be your best answer
Blue screen tends to corrupt any database that is running.

You need to find out what caused your blue screen.
Check disk, Clean disk, Defrag disk. will help.

---

http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3549065

interesting that a few people couldnt download this file

---

I've already reformatted and reinstalled a few weeks ago.

> > can anyone tell me what happened here?
> > http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3819102
> >
> > 2 others get credit, i got 0.00 It's invalid? i'm wondering if this is
> because
> > maybe windows blue screened at the wrong moment?
> >
> Mighty Mik
>
> I would suspect that the Blue screen would be your best answer
> Blue screen tends to corrupt any database that is running.
>
> You need to find out what caused your blue screen.
> Check disk, Clean disk, Defrag disk. will help.
>

> http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3549065
>
> interesting that a few people couldnt download this file
>

All users got download-errors on this wu, even they marked as "success". This wu was generated 27. October, right before the disk-crash, so most likely got corrupted on disk.

> I've already reformatted and reinstalled a few weeks ago.
> > Mighty Mik
> >
> > I would suspect that the Blue screen would be your best answer
> > Blue screen tends to corrupt any database that is running.
> >
> > You need to find out what caused your blue screen.
> > Check disk, Clean disk, Defrag disk. will help.
> >
>
Mighty Mik
I would suspect a hardware failure or conflict....

---

> can anyone tell me what happened here?
> http://setiweb.ssl.berkeley.edu/workunit.php?wuid=3819102
>
> 2 others get credit, i got 0.00 It's invalid? i'm wondering if this is because
> maybe windows blue screened at the wrong moment?
>


Any number of things can cause an invalid. I have quite a few... and no blue screens.

As to what caused the blue screen, possibilites are limitless. Could be you have something made for SP3 installed on a SP1 machine, possibly some update from AMD.

Could be a program your running, possibly even BOINC... or just some type of corruption in your RAM/page file that caused blue screen.

Any chance of retrieving the error name from the blue screen? I know MS gives you the option to report errors, if you did it like you should the error will be on microsofts site somewhere. If not reported, the file it was supposed to send might still be on your disk. Lemme know I'll help ya find it.

---

I would too...except i have dual boot, and it doesn't seem to happen under LINUX.

as far as error info...if ya blink, ya miss it, it's that fast. The theory is...the CPU is a little flakey. Windows does something to it it doesn't like, but if i run Linux, it likes it better.I'm saving for a new Mac of some kind, after that, i'm looking into watercooling my PC. I don't want to build a 64bit game monster with a $400+ vid card and a 500W power supply right now.

> > I've already reformatted and reinstalled a few weeks ago.
> > > Mighty Mik
> > >
> > > I would suspect that the Blue screen would be your best answer
> > > Blue screen tends to corrupt any database that is running.
> > >
> > > You need to find out what caused your blue screen.
> > > Check disk, Clean disk, Defrag disk. will help.
> > >
> >
> Mighty Mik
> I would suspect a hardware failure or conflict....
>