Check this box out? Do you think its has a problem? Can it be reported to the owner? Or perhaps to the Seti staff?
Bad computer

Looks like someone has found a way to cheat in my humble opinion


Result ID 75502418
Name 18au03aa.8872.9250.198600.28_3
Workunit 18172028
Created 18 Jun 2005 7:57:18 UTC
Sent 19 Jun 2005 15:40:25 UTC
Received 20 Jun 2005 1:15:27 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x0)
Computer ID 987127
Report deadline 3 Jul 2005 15:40:25 UTC
CPU time 0
stderr out 4.68





Validate state Valid
Claimed credit 0
Granted credit 13.2689230648062
application version 4.07


Has quite a few of these valid, zero time, zero credit WUs, but is still getting credit from the other crunchers, because his are valid, just seems wrong..... Running CC 4.68 ?????????

---

Semper Eadem
So long Paul, it has been a hell of a ride.

Park your ego's, fire up the computers, Science YES, Credits No.

LMAO! Amazing!

---

[font='fixedsys,courier']There is a whiff of fishy in the air... reeks of cheat.[/font]

---

Looks like someone has found a way to cheat in my humble opinion


Result ID 75502418
Name 18au03aa.8872.9250.198600.28_3
Workunit 18172028
Created 18 Jun 2005 7:57:18 UTC
Sent 19 Jun 2005 15:40:25 UTC
Received 20 Jun 2005 1:15:27 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x0)
Computer ID 987127
Report deadline 3 Jul 2005 15:40:25 UTC
CPU time 0
stderr out 4.68





Validate state Valid
Claimed credit 0
Granted credit 13.2689230648062
application version 4.07


Has quite a few of these valid, zero time, zero credit WUs, but is still getting credit from the other crunchers, because his are valid, just seems wrong..... Running CC 4.68 ?????????

I wonder if we put ROM 'S name in here if he would have someone check it out and maybe straighten us out if needed?
I also did not know there was a version 4.68

---

Looks like someone has found a way to cheat in my humble opinion

No, his results are valid. That is no way to cheat. I believe he uses self compiled BOINC and SETI and there is bug in reporting the CPU time. This give the problem with the claimed credit = 0.

Sounds to me as if the way to use at least 3 valid results for computing granted credit and to throw away the upper and lower end and only using the average of the middle results works.

I reported it to the staff.

---

Greetings from Bremen/Germany
Jens Seidler (TheBigJens)

Looks like someone has found a way to cheat in my humble opinion

No, his results are valid. That is no way to cheat. I believe he uses self compiled BOINC and SETI and there is bug in reporting the CPU time. This give the problem with the claimed credit = 0.

Sounds to me as if the way to use at least 3 valid results for computing granted credit and to throw away the upper and lower end and only using the average of the middle results works.

Okay that makes sense, but shouldn't Berkeley be slowly shutting these computers down? Shouldn't you guys be giving these computers fewer and fewer units until the problem is addressed? I mean you DO have our email addresses.

---

Okay that makes sense, but shouldn't Berkeley be slowly shutting these computers down? Shouldn't you guys be giving these computers fewer and fewer units until the problem is addressed? I mean you DO have our email addresses.

Clear, at least the user will get a mail. Catching the error automaticly may need some more thoughts.

---

Greetings from Bremen/Germany
Jens Seidler (TheBigJens)

I know what the problem is - benchmark gone wrong, this started happening to my linux box 2 days ago, but no one has even responded to my q&a 2 days ago to help diagnose the problem. I refer you to here for my benchmark problem. It looks like a boinc problem. I am using the offical boinc cc 4.43 and seti apps for this box. Reboot does nothing, have not reinstalled as I would like to debug this problem with some help please.

Bones

Things do look strange on your Linux machine. Did you notice that it has 0 CPU's and that it has no benchmark times?

---

[font='fixedsys,courier']Sorry about that - It looked very cheat-like.

IIRC BOINC will default to 1MIPS when the benchmarks haven't been run - not a flat 0.

Can you try upgrading to Sarge? Or using GCC?[/font]

---

I also did not know there was a version 4.68

4.68 is the recommended version for the CPDN Alpha project.

@Neo, the benchmarks have run with the results now showing zero mips and 0 cpu, this happened 2 days ago. Before that the pc previously had correct benchmarks of, from memory, around 400 double mips and 800 integer mips (which is lower than this pc should report but I am aware of the bug with the linux boinc version causing this error).

I don't think my linux version is the problem as it was all working fine until 2 days ago (I haven't changed anything on the box, it just cruches s@h 24/7). If you go back through the previous results (

---

Is there a limit to characters in a post, I just lost have my post?

continuing, go back prior to 18 june and all claimed credits are normal. A manual bencmark produces

number of cpu = 0
nan double mips
nan intger mips

hence why I now claim 0 credits as these numbers are used in the credit calculation

---

One of my PCs grant about 28 credits and gets 0.... while the otehr PCs that crunched it also got about 30

---

One of my PCs grant about 28 credits and gets 0.... while the otehr PCs that crunched it also got about 30

You returned the results too late... one of them 3 hours too late and
the other one 1 day too late...

*Hint: 2 week deadline ;)

---

[font='fixedsys,courier']Hmm... the only things I can think that either the benchmarks were run in the background by the BOINC client, and the system was under heavy workload, or new hardware was added that the OS and/or BOINC couldn't handle...

Either way, I'm stumped.[/font]

---

Actually seeing as the Computer is FreeBSD I bet the downloaded work units are aborting with a Segment Violation.

It took me some fiddling before I actually got my FreeBSD Machine Crunching. I agree that the owner of this computer needs to be notified that it's failing big time.

---

I'd rather speak my mind because it hurts too much to bite my tongue.

American Spirit BBQ Proudly Serving those that courageously defend freedom.

Hmmm...I don't think its really "failing", the results are validating correctly with a success outcome. My linux box is doing a similiar thing (and still gets successful results) but it can stay like that until I finally d/l sarge and reinstall everything......again

At least here the median result is chosen for granted credit, unlike predictor where the same thing has happened and all crunchers for that wu get 0.00. See this predictor thread1759

[font='fixedsys,courier']Ow!

Methinks the Devs've gotta fix that...[/font]

---