computation errors


log in

Advanced search

Message boards : Number crunching : computation errors

Previous · 1 · 2
Author Message
Mox
Send message
Joined: 8 Apr 09
Posts: 31
Credit: 372,650
RAC: 0
Message 963936 - Posted: 16 Jan 2010, 22:05:26 UTC - in response to Message 963927.

i'm sorry, but where can't i get this rescheduler and so on?

Profile Raistmer
Volunteer developer
Volunteer tester
Avatar
Send message
Joined: 16 Jun 01
Posts: 3644
Credit: 49,364,786
RAC: 29,403
Russia
Message 964046 - Posted: 17 Jan 2010, 6:59:24 UTC - in response to Message 963936.

i'm sorry, but where can't i get this rescheduler and so on?

@Lunatics of course ;)
http://lunatics.kwsn.net/gpu-crunching/gpu-cpu-rebranding-script.0.html

Rasputin
Volunteer tester
Send message
Joined: 13 Jun 02
Posts: 1764
Credit: 6,132,221
RAC: 0
Russia
Message 964051 - Posted: 17 Jan 2010, 8:50:47 UTC

I had to suspend new work download because almost everything I've downloaded in the last two days have been VLAR's. 677 of them to be exact!! I don't think I'll make the deadline on all of them.

I'll run three cores for awhile to help clear them out. meanwhile, I'll most likely run out of CUDA tasks for my three GPU's.

Why are we getting so many VLAR's lately?

Richard HaselgroveProject donor
Volunteer tester
Send message
Joined: 4 Jul 99
Posts: 8813
Credit: 53,489,047
RAC: 45,413
United Kingdom
Message 964053 - Posted: 17 Jan 2010, 10:27:06 UTC - in response to Message 964051.

I had to suspend new work download because almost everything I've downloaded in the last two days have been VLAR's. 677 of them to be exact!! I don't think I'll make the deadline on all of them.

I'll run three cores for awhile to help clear them out. meanwhile, I'll most likely run out of CUDA tasks for my three GPU's.

Why are we getting so many VLAR's lately?

SETI doesn't have the funds to buy and run a radio telescope all of its own. Instead, it goes along for a free ride, recording whatever the Arecibo radio telescope happens to be pointing at as part of somebody else's research.

You can see which research group is using the telescope at any particular time by searching the Arecibo Observatory Telescope Schedule. There seem to be a number of VLARs recorded on 2/3/4 January 2007 in my cache at the moment: I guess those would be from A2196: HI Emission Profiles in AGN Hosts: A New Strategy to Test the Black Hole-Host Galaxy Paradigm.

We've been peering into black holes. Perhaps not the most likely place to find ET, but beggars can't be choosers.

Profile Frosted
Avatar
Send message
Joined: 11 Jul 99
Posts: 83
Credit: 3,898,641
RAC: 0
Canada
Message 964075 - Posted: 17 Jan 2010, 12:10:06 UTC - in response to Message 964046.
Last modified: 17 Jan 2010, 12:10:39 UTC

i'm sorry, but where can't i get this rescheduler and so on?

@Lunatics of course ;)
http://lunatics.kwsn.net/gpu-crunching/gpu-cpu-rebranding-script.0.html


I can't get that link to work.
Tyr this
http://atsplus.ru/boinc/ReSchedule1.9.rar
____________

Rasputin
Volunteer tester
Send message
Joined: 13 Jun 02
Posts: 1764
Credit: 6,132,221
RAC: 0
Russia
Message 964120 - Posted: 17 Jan 2010, 15:41:53 UTC - in response to Message 964053.

I had to suspend new work download because almost everything I've downloaded in the last two days have been VLAR's. 677 of them to be exact!! I don't think I'll make the deadline on all of them.

I'll run three cores for awhile to help clear them out. meanwhile, I'll most likely run out of CUDA tasks for my three GPU's.

Why are we getting so many VLAR's lately?

SETI doesn't have the funds to buy and run a radio telescope all of its own. Instead, it goes along for a free ride, recording whatever the Arecibo radio telescope happens to be pointing at as part of somebody else's research.

You can see which research group is using the telescope at any particular time by searching the Arecibo Observatory Telescope Schedule. There seem to be a number of VLARs recorded on 2/3/4 January 2007 in my cache at the moment: I guess those would be from A2196: HI Emission Profiles in AGN Hosts: A New Strategy to Test the Black Hole-Host Galaxy Paradigm.

We've been peering into black holes. Perhaps not the most likely place to find ET, but beggars can't be choosers.


I knew all of that except about the black holes. It does sound like an unlikely place to find ET. Never know though, there could be a planet that has an orbit between here and the black hole that crosses paths with it.

Your other post here: http://setiathome.berkeley.edu/forum_thread.php?id=58428&nowrap=true#963852 answered something else I wanted to know... i.e. What exactly are these VLAR wu's that we're forced to reschedule or kill.

Thanks Richard! Very easy to understand explanations even for someone like me with a very limited knowledge of astronomy. :)

Profile perryjay
Volunteer tester
Avatar
Send message
Joined: 20 Aug 02
Posts: 3377
Credit: 16,370,608
RAC: 9,057
United States
Message 964131 - Posted: 17 Jan 2010, 16:38:50 UTC

But maybe that's what happened to ET,..........



Silly, everyone knows ET use black holes to travel from one universe to another! :-)
____________


PROUD MEMBER OF Team Starfire World BOINC

Profile Phil J Taylor
Avatar
Send message
Joined: 26 Dec 09
Posts: 96
Credit: 603,521
RAC: 0
United States
Message 964139 - Posted: 17 Jan 2010, 17:26:03 UTC

Maybe this Workunit might help. Four machines running v6.09 (cuda23) failed with the same error message:

cudaAcc_find_triplets erroneously found a triplet twice in find_triplets_kernel File: c:/sw/gpgpu/seti/seti_boinc/client/cuda/cudaAcc_pulsefind.cu Line: 236


Two machines running v6.03 were verified and earned credit for the same WU.

One very helpful anonymous platform included debugging info in their result.

Profile Gundolf Jahn
Send message
Joined: 19 Sep 00
Posts: 3184
Credit: 361,286
RAC: 37
Germany
Message 964150 - Posted: 17 Jan 2010, 18:08:04 UTC - in response to Message 964136.

Silly, everyone knows ET use black holes to travel from one universe to another! :-)

I thought they were wormholes.....

Yes, but you need a black hole as energy source for an intergalactic wormhole ;-)

Gruß,
Gundolf
____________
Computer sind nicht alles im Leben. (Kleiner Scherz)

SETI@home classic workunits 3,758
SETI@home classic CPU time 66,520 hours

Profile Davissimo
Avatar
Send message
Joined: 28 Jul 08
Posts: 27
Credit: 605,971
RAC: 640
United States
Message 964425 - Posted: 18 Jan 2010, 19:13:24 UTC

A large number of tasks have been failing on my machine and I have been asking for help from my team. Team lead Chris Marion pointed me to your threads. Our discussions are on the Alien Beacon Tech Questions. But in a nutshell I started using Lunatics optimizer and after just a couple days the problems began. The error messages are
1/14/2010 7:05:35 PM SETI@home Starting 16fe07aa.26631.11115.13.10.227_1
1/14/2010 7:05:38 PM SETI@home [error] Process creation failed:
1/14/2010 7:05:40 PM SETI@home Computation for task 16fe07aa.26631.11115.13.10.227_1 finished
1/14/2010 7:05:40 PM SETI@home Output file 16fe07aa.26631.11115.13.10.227_1_0 for task 16fe07aa.26631.11115.13.10.227_1 absent

It also showed the application as "Anonymous Platform". About 60 tasks have failed so I've suspended SETI for the moment hoping not to do any more damage, if damage is indeed done. The reading of this thread and a few others has been informative but no resolution is seen. I do not have a GPU so the VLAR is likely not a problem for me. I could have done (somehow) a bad install of the optimizer. Chis thought that I just got a bad batch. Or there's that Windows update to which I haven't done a restart. Should I consider backing out the optimizer? It's not the end of the world of course but I would like my little machine to be productive. Your thoughts?
____________
All computers idle at the same speed.

Profile HAL9000
Volunteer tester
Avatar
Send message
Joined: 11 Sep 99
Posts: 4664
Credit: 123,560,374
RAC: 98,553
United States
Message 964427 - Posted: 18 Jan 2010, 19:35:17 UTC - in response to Message 964425.

A large number of tasks have been failing on my machine and I have been asking for help from my team. Team lead Chris Marion pointed me to your threads. Our discussions are on the Alien Beacon Tech Questions. But in a nutshell I started using Lunatics optimizer and after just a couple days the problems began. The error messages are
1/14/2010 7:05:35 PM SETI@home Starting 16fe07aa.26631.11115.13.10.227_1
1/14/2010 7:05:38 PM SETI@home [error] Process creation failed:
1/14/2010 7:05:40 PM SETI@home Computation for task 16fe07aa.26631.11115.13.10.227_1 finished
1/14/2010 7:05:40 PM SETI@home Output file 16fe07aa.26631.11115.13.10.227_1_0 for task 16fe07aa.26631.11115.13.10.227_1 absent

It also showed the application as "Anonymous Platform". About 60 tasks have failed so I've suspended SETI for the moment hoping not to do any more damage, if damage is indeed done. The reading of this thread and a few others has been informative but no resolution is seen. I do not have a GPU so the VLAR is likely not a problem for me. I could have done (somehow) a bad install of the optimizer. Chis thought that I just got a bad batch. Or there's that Windows update to which I haven't done a restart. Should I consider backing out the optimizer? It's not the end of the world of course but I would like my little machine to be productive. Your thoughts?


Did you use the Win32 Lunatics' Unified Installer v0.2 or download a single version like "4 - Windows 32 bit AK v8.0 SSE (Pentium 3)". If you are using the incorrect one you will get lots of issues.
You might wish to read the sticky topic AK V8 ported release ap. issues, install, questions etc.
____________
SETI@home classic workunits: 93,865 CPU time: 863,447 hours

Join the BP6/VP6 User Group today!

Profile Gundolf Jahn
Send message
Joined: 19 Sep 00
Posts: 3184
Credit: 361,286
RAC: 37
Germany
Message 964433 - Posted: 18 Jan 2010, 19:51:14 UTC - in response to Message 964425.
Last modified: 18 Jan 2010, 19:52:59 UTC

It also showed the application as "Anonymous Platform".

If you are running a 3rd-party application (Lunatics), that is using Anonymous Platform (AP), not to confuse with AstroPulse. ;-)

...
Or there's that Windows update to which I haven't done a restart.

A reboot might actually help! See the BOINC FAQ Service:
ERR_RESULT_START -185 This is an error that will occur when: - if BOINC couldn't start the application - if files are missing - catch other error returns - on nonzero exit or signal - if exceeded resource limit - catch-all for resume/start errors Rebooting your machine will most of the times fix these.


Should I consider backing out the optimizer?

Perhaps (if a repair install of BOINC doesn't help) a downgrade of BOINC does help. See Message 948016 [edit]and following[/edit].

Gruß,
Gundolf
____________
Computer sind nicht alles im Leben. (Kleiner Scherz)

SETI@home classic workunits 3,758
SETI@home classic CPU time 66,520 hours

Profile Davissimo
Avatar
Send message
Joined: 28 Jul 08
Posts: 27
Credit: 605,971
RAC: 640
United States
Message 965262 - Posted: 22 Jan 2010, 19:30:19 UTC

The problems continue. A restart did not do any good. An attempt to re-install an older version of Boinc resulted in an error message. A re-install of the current version seemed to go okay but SETI tasks again failed. Then an uninstall was done on Lunatics but still SETI failed and showed "Anonymous Platform" as the application. It seems that an uninstall of Boinc is the next step, but I hesitate. Will it damage my stats or will those be recovered from the web sites?
____________
All computers idle at the same speed.

Profile ignorance is no excuse
Avatar
Send message
Joined: 4 Oct 00
Posts: 9529
Credit: 44,433,321
RAC: 0
Korea, North
Message 965272 - Posted: 22 Jan 2010, 20:06:19 UTC - in response to Message 965262.

your error message is -185
answers for -185 from boinc faq
ERR_RESULT_START -185

This is an error that will occur when:
- if BOINC couldn't start the application
- if files are missing
- catch other error returns
- on nonzero exit or signal
- if exceeded resource limit
- catch-all for resume/start errors

Rebooting your machine will most of the times fix these.

have you upgraded boinc recently? perhaps the files are in the wrong place?

since you aren't even able to start the WU's I would reset the project and if that doesnt help detach/reattach to the project
____________
In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope

End terrorism by building a school

Profile perryjay
Volunteer tester
Avatar
Send message
Joined: 20 Aug 02
Posts: 3377
Credit: 16,370,608
RAC: 9,057
United States
Message 965273 - Posted: 22 Jan 2010, 20:06:34 UTC - in response to Message 965262.

Well, it looks like your stats are already being damaged. If you uninstall BOINC just be sure you use the same email address and username and password when you reinstall. Your stats and everything will be ok then. The FAQs have a very good tutorial on how to completely remove BOINC here.. http://boincfaq.mundayweb.com/index.php?language=1&view=490&sessionID=55885c676e3392a870137a93646d2956 Follow Jord's directions, then get everything back up and running stock first. Once you are sure stock is turning in good work we can help you get optimized again.
____________


PROUD MEMBER OF Team Starfire World BOINC

Profile Davissimo
Avatar
Send message
Joined: 28 Jul 08
Posts: 27
Credit: 605,971
RAC: 640
United States
Message 965542 - Posted: 23 Jan 2010, 20:43:36 UTC

Having followed the tutorial pointed to by perryjay, BOINC was successfully uninstalled then easily reinstalled. So far all projects tried have good running tasks including SETI. I want to thank everyone for their help. Perhaps I may ask for advice later in getting a good install of Lunatics but for the moment I'm just going to sit back and enjoy watching SETI run.
____________
All computers idle at the same speed.

Previous · 1 · 2

Message boards : Number crunching : computation errors

Copyright © 2014 University of California