cooldudeof02

Welcome to Seti BOINC

I wish that I could see the post you are referening to... Yes as a result of a couple of "runaway machines" the problem where they were burning in that case thousands of results was fixed... Scheduler shut them off to get their attention. Now we need to see on of machine ID's to help determine the problem.

Just had 42 computation errors in a row, 0 process time. Now can't download units because I have already had too many today! Never had any problems until now. If the enhanced Seti isn't stable why release it?

Seti enhanced has been quite stable for me in the month that I've beta tested it. Perhaps if more people had beta tested it, these errors would've been located and fixed whilst it was still in the beta phase!

So Yes, I am also finding Seti Enhanced Stable... but then if we look at the number of computers in use with Seti Beta (1,750 active), that equals about 0.0052 percent of the active hosts here at Seti BOINC..

Edit for the percentage

Pappa

---

Please consider a Donation to the Seti Project.

Yeah, that one's got the jpeg bug bad.

Eric

http://setiathome.berkeley.edu/show_host_detail.php?hostid=2376252
This computer can crunch the 4.18 units just fine. About 2/3 of the 5.12 units are erroring out in 1-2 minutes.
The error code is the same in each case:
5/7/2006 5:43:19 PM|SETI@home|Starting task 06ja99ab.28689.18867.720172.3.68_1 using setiathome_enhanced version 512
5/7/2006 5:45:11 PM|SETI@home|Unrecoverable error for result 06ja99ab.28689.18867.720172.3.68_1 ( - exit code -1073741819 (0xc0000005))

---

@SETIEric@qoto.org (Mastodon)

That's why i can't understand why Eric relased it to the public.

Ever wonder why the word "dead" is in the word deadline. :(

Eric

---

@SETIEric@qoto.org (Mastodon)

???, What does that mean and what do I do to fix it?

Yeah, that one's got the jpeg bug bad.

Eric

http://setiathome.berkeley.edu/show_host_detail.php?hostid=2376252
This computer can crunch the 4.18 units just fine. About 2/3 of the 5.12 units are erroring out in 1-2 minutes.
The error code is the same in each case:
5/7/2006 5:43:19 PM|SETI@home|Starting task 06ja99ab.28689.18867.720172.3.68_1 using setiathome_enhanced version 512
5/7/2006 5:45:11 PM|SETI@home|Unrecoverable error for result 06ja99ab.28689.18867.720172.3.68_1 ( - exit code -1073741819 (0xc0000005))

---

---

The Postman

It means in the projects\\setiathome.berkeley.edu folder that the file "seti512.jpg" has a problem... It is what is causing your problems... At the moment I can not find the link to the fanout where you can stop BOINC and download the file and save it to the folder... That will correct the problem... Otherwise, it gets to be a bit tougher... Hopefully someone can give you the correct link to download the file... Or copy from another computer...

Pappa

???, What does that mean and what do I do to fix it?

Yeah, that one's got the jpeg bug bad.

Eric

http://setiathome.berkeley.edu/show_host_detail.php?hostid=2376252
This computer can crunch the 4.18 units just fine. About 2/3 of the 5.12 units are erroring out in 1-2 minutes.
The error code is the same in each case:
5/7/2006 5:43:19 PM|SETI@home|Starting task 06ja99ab.28689.18867.720172.3.68_1 using setiathome_enhanced version 512
5/7/2006 5:45:11 PM|SETI@home|Unrecoverable error for result 06ja99ab.28689.18867.720172.3.68_1 ( - exit code -1073741819 (0xc0000005))

---

Please consider a Donation to the Seti Project.

In reply to Ageless (a Forum Moderator) from this thread, here is the info requested:

WU Crashed - Result ID 319803702
http://setiathome.berkeley.edu/result.php?resultid=319803702

OS:
Windows XP Professional Version 2002 - Service Pack 2

BOINC Version:
Version 5.2.13

Video Card:
ATI Mobility Radeon 9600
Driver Version 6.14.10.6404
Driver Date 11/13/2003
Driver Provider ATI Technologies Inc.
(note: yea, I know; driver is a bit old, BUT I don't play games. lol - But the source problem was that the work unit was not even progressing (stuck at 6.50%) after nearly half a day (12 hours) ...)
As for Direct X, how do we look that up? I cannot find it anywhere. All I know is that EverQuest (which always requires the latest versions) works.

And this type of "crash" reminds me; I need to return my movie rental, "Crash" sometime this week! hahaha

Regarding the over abundance of -9 overflow results, these tend to come in batches. It's usually not so much of a problem when we are running 6 splitters at once because 5/6th of the work units will be coming from other tapes.

I've upped the number of enhanced splitters to 2 so this should become less of a problem.

Eric

I am probably wrong here but..........

I carry a 2 day cache on 4 computers. Thats a total cache size of 102 at the moment and of these 21 are for version 4.18 and 81 are seti-enhanced version 5.12. This seems backwards to the procedure described for the transition where the number of 5.12 would be incremented during a time period of one month. Am I wrong here or do I have a weird cache?

---

Boinc....Boinc....Boinc....Boinc....

Eric or Pappa.

Would copying the jpeg from my beta folder to the seti main folder work?

Are they the same file?

---

Strange. Eric is using his Computer since 1901!!!! Is he ET. Can we stop crunching?

---

Peter

I am not sure that it the same... the best location to retrieve corrupt files from would be:

http://boinc2.ssl.berkeley.edu/sah/download_fanout/
seti512.jpg

Stop the BOINC manager/service, then download (right click and Save As in Windows) the seti512.jpg to the projects\\setiathome.berkeley.edu folder...

Eric or Pappa.

Would copying the jpeg from my beta folder to the seti main folder work?

Are they the same file?

---

Please consider a Donation to the Seti Project.

Since the upgrade to version 5.12 of SETI client, I have had 1 invalid work unit and several work-units where within the Boinc Manager's 'Work' window, the CPU time stays at --- and the Process column stayed at 0.00% while the SETI client is actively working on the unit.

Pressing the 'Show graphics' button of the 5.2.13 version of Boinc Manager shows an overall % done that is increasing and appears to be reasonable given the amount of time that the client has spent working on the work unit, but the 'CPU time' field of that graphic display stays at 0.00.

I have noticed that if I restart the Boinc Manager, the same work-unit which had been worked on previously will now start showing both the 'CPU time' and 'Progress' columns increase in value under the 'Work' window of the Boinc Manager. However, the unit appears to be being completely re-evaluated rather than continuing it's processing from where it last left off.

---

http://setiweb.ssl.berkeley.edu/workunit.php?wuid=76873327

cpu time = 0

CPU time 0
stderr out <core_client_version>4.32</core_client_version>
<stderr_txt>
ar=0.426499 NumCfft=72527 NumGauss= 459901078 NumPulse= 88118656383 NumTriplet= 7643560755200
No heartbeat from core client for 31 sec - exiting
ar=0.426499 NumCfft=72527 NumGauss= 459901078 NumPulse= 88118656383 NumTriplet= 7643560755200

</stderr_txt>


Validate state Valid
Claimed credit 0
------------------------------
Barton 2500+ - Win98SE - CC 4.32

---

Do you want to get banned for 31 years and your account & credits deleted at a Boinc project ? Predictor@home is your best choice.

Since the upgrade to version 5.12 of SETI client, I have had 1 invalid work unit and several work-units where within the Boinc Manager's 'Work' window, the CPU time stays at --- and the Process column stayed at 0.00% while the SETI client is actively working on the unit.

Pressing the 'Show graphics' button of the 5.2.13 version of Boinc Manager shows an overall % done that is increasing and appears to be reasonable given the amount of time that the client has spent working on the work unit, but the 'CPU time' field of that graphic display stays at 0.00.

I have noticed that if I restart the Boinc Manager, the same work-unit which had been worked on previously will now start showing both the 'CPU time' and 'Progress' columns increase in value under the 'Work' window of the Boinc Manager. However, the unit appears to be being completely re-evaluated rather than continuing it's processing from where it last left off.

It's a known problem afflicting Win9X based systems, see the Beta Test thread "No checkpoints, no progress, but completes OK". Eric noted this timer problem as something remaining to be fixed when he announced that enhanced was going mainstream. It seems to happen fairly frequently on some systems, very rarely if at all on some other Win9X systems.

If you can simply let it run to completion, that's best. OTOH, if your system is doing more than one project you'd have to guess the likelihood of the WU being preempted and having to start over.

                                                               Joe

Not sure if this is an error or not, but i have recdeived a HUGE, and i mean HUGE WU.

My system has been sat pretty much idle working on seti WU's for the last month, and i thought after a few days away from it i might check and see if it was still running.
This message is where the wu started
06/05/2006 19:09:34|SETI@home|Starting result 06ja99ab.28689.3411.657686.3.20_2 using setiathome_enhanced version 512
It is still processing now. currently at 50 hours CPU time, 3.42% complete, and approx 53 hours to completion?????
anyone else got this WU, and getting the same kind of report?

06ja99ab.28689.3411.657686.3.20_2

It's a known problem afflicting Win9X based systems, see the Beta Test thread "No checkpoints, no progress, but completes OK". Eric noted this timer problem as something remaining to be fixed when he announced that enhanced was going mainstream. It seems to happen fairly frequently on some systems, very rarely if at all on some other Win9X systems.

I've also discovered some thread synchronization problems in the BOINC api that I will try to resolve at the same time that I fix this. The 0xc0000005 faults while reading the JPEG files appear to be due to in interaction with how jpeglib handles error and the BOINC exception handler.

One advantage of having lots of examples of errors is that lots of examples can help when figuring things out.

Eric

---

@SETIEric@qoto.org (Mastodon)

Eric or Pappa.

Would copying the jpeg from my beta folder to the seti main folder work?

Are they the same file?

Yep, they should be identical.

Eric

---

@SETIEric@qoto.org (Mastodon)

http://setiathome.berkeley.edu/result.php?resultid=320467931
http://setiathome.berkeley.edu/result.php?resultid=317511265

---

Since the upgrade to version 5.12 of SETI client, I have had 1 invalid work unit and several work-units where within the Boinc Manager's 'Work' window, the CPU time stays at --- and the Process column stayed at 0.00% while the SETI client is actively working on the unit.

Pressing the 'Show graphics' button of the 5.2.13 version of Boinc Manager shows an overall % done that is increasing and appears to be reasonable given the amount of time that the client has spent working on the work unit, but the 'CPU time' field of that graphic display stays at 0.00.

I have noticed that if I restart the Boinc Manager, the same work-unit which had been worked on previously will now start showing both the 'CPU time' and 'Progress' columns increase in value under the 'Work' window of the Boinc Manager. However, the unit appears to be being completely re-evaluated rather than continuing it's processing from where it last left off.

It's a known problem afflicting Win9X based systems, see the Beta Test thread "No checkpoints, no progress, but completes OK". Eric noted this timer problem as something remaining to be fixed when he announced that enhanced was going mainstream. It seems to happen fairly frequently on some systems, very rarely if at all on some other Win9X systems.

If you can simply let it run to completion, that's best. OTOH, if your system is doing more than one project you'd have to guess the likelihood of the WU being preempted and having to start over.

                                                               Joe

Umm Joe, this is another one that is not win98 but close:
GenuineIntel
Intel(R) Pentium(R) 4 CPU 3.20GHz Microsoft Windows Millennium
, (04.90.3000.00)

As I have stated over at beta this is not just win98 issue!!

---

98SE XP2500+ @ 2.1 GHz Boinc v5.8.8

And God said"Let there be light."But then the program crashed because he was trying to access the 'light' property of a NULL universe pointer.

Since the upgrade to version 5.12 of SETI client, I have had 1 invalid work unit and several work-units where within the Boinc Manager's 'Work' window, the CPU time stays at --- and the Process column stayed at 0.00% while the SETI client is actively working on the unit.

Pressing the 'Show graphics' button of the 5.2.13 version of Boinc Manager shows an overall % done that is increasing and appears to be reasonable given the amount of time that the client has spent working on the work unit, but the 'CPU time' field of that graphic display stays at 0.00.

I have noticed that if I restart the Boinc Manager, the same work-unit which had been worked on previously will now start showing both the 'CPU time' and 'Progress' columns increase in value under the 'Work' window of the Boinc Manager. However, the unit appears to be being completely re-evaluated rather than continuing it's processing from where it last left off.

It's a known problem afflicting Win9X based systems, see the Beta Test thread "No checkpoints, no progress, but completes OK". Eric noted this timer problem as something remaining to be fixed when he announced that enhanced was going mainstream. It seems to happen fairly frequently on some systems, very rarely if at all on some other Win9X systems.

If you can simply let it run to completion, that's best. OTOH, if your system is doing more than one project you'd have to guess the likelihood of the WU being preempted and having to start over.

                                                               Joe

Umm Joe, this is another one that is not win98 but close:
GenuineIntel
Intel(R) Pentium(R) 4 CPU 3.20GHz Microsoft Windows Millennium
, (04.90.3000.00)

As I have stated over at beta this is not just win98 issue!!

I beleive that "Win 9x" (note "x") means all the Windows releases that are/were based on the Win 95 core code. These are Win 95, 98, 98r2, and Me. (as opposed to the "NT" group, IE Win Nt, 2k, & XP...) I'm not sure what "longhorn" is based on, if it isn't a complete re-write...

---

.

Hello, from Albany, CA!...