Computation Error

Message boards : Number crunching : Computation Error

To post messages, you must log in.

AuthorMessage
Profile Tom Lee

Send message
Joined: 10 Dec 05
Posts: 5
Credit: 129,728
RAC: 0
Message 11098 - Posted: 21 Feb 2006, 11:36:08 UTC

I'm having more and more wu that end-up with a computation error status. I've had 15 hours worth in the last 24. I'm not sure what this means. My computer can't add, subtract, and divide, or it's not following program instructions? Do I have a rogue CPU? Could someone enlighten me - I can't find mention of it on the message boards. Thanks.
ID: 11098 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Moderator9
Volunteer moderator

Send message
Joined: 22 Jan 06
Posts: 1014
Credit: 0
RAC: 0
Message 11102 - Posted: 21 Feb 2006, 12:18:57 UTC - in response to Message 11098.  

I'm having more and more wu that end-up with a computation error status. I've had 15 hours worth in the last 24. I'm not sure what this means. My computer can't add, subtract, and divide, or it's not following program instructions? Do I have a rogue CPU? Could someone enlighten me - I can't find mention of it on the message boards. Thanks.



Check out this FAQ in the FAQs thread

Moderator9
ROSETTA@home FAQ
Moderator Contact
ID: 11102 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Tom Lee

Send message
Joined: 10 Dec 05
Posts: 5
Credit: 129,728
RAC: 0
Message 11150 - Posted: 21 Feb 2006, 20:21:29 UTC - in response to Message 11149.  

I'm having more and more wu that end-up with a computation error status. I've had 15 hours worth in the last 24. I'm not sure what this means. My computer can't add, subtract, and divide, or it's not following program instructions? Do I have a rogue CPU? Could someone enlighten me - I can't find mention of it on the message boards. Thanks.



Check out this FAQ in the FAQs thread


The status specifically says "computation error". I can't find this in the FAQ. So what's happened when I get this message?
ID: 11150 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Moderator9
Volunteer moderator

Send message
Joined: 22 Jan 06
Posts: 1014
Credit: 0
RAC: 0
Message 11168 - Posted: 22 Feb 2006, 0:45:02 UTC - in response to Message 11150.  
Last modified: 22 Feb 2006, 0:50:42 UTC

I'm having more and more wu that end-up with a computation error status. I've had 15 hours worth in the last 24. I'm not sure what this means. My computer can't add, subtract, and divide, or it's not following program instructions? Do I have a rogue CPU? Could someone enlighten me - I can't find mention of it on the message boards. Thanks.



Check out this FAQ in the FAQs thread


The status specifically says "computation error". I can't find this in the FAQ. So what's happened when I get this message?



"Computation error" in the work display of BOINC is a catch all term used for a number of different problems.

If you look at the workunit in you stats page, you will find the "exception" errors in the errors reported for the WUs. What I am seeing is that with the exception of one of your errors, other systems have also failed to complete the same work units. This tells me that it may be a work unit issue. You really do not have that many errors. and the one that is unique to your system may also turn out to be work unit related but the other system that it was issued to has not reported back yet.

Based on what I saw on your stats page, I refered you to that particular FAQ. It is possible that there was some transient probelm on your system that has caused the one error that does not look like a WU issue.

Your system does seem marginal on memory, you might try increasing your swap time if you are running more than one project and see if that helps.

Moderator9
ROSETTA@home FAQ
Moderator Contact
ID: 11168 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Morphy375
Avatar

Send message
Joined: 2 Nov 05
Posts: 86
Credit: 1,629,758
RAC: 0
Message 11212 - Posted: 22 Feb 2006, 8:43:52 UTC

I have several machines with 256MB running Rosetta without any problems.....
Teddies....
ID: 11212 · Rating: 1 · rate: Rate + / Rate - Report as offensive    Reply Quote
Moderator9
Volunteer moderator

Send message
Joined: 22 Jan 06
Posts: 1014
Credit: 0
RAC: 0
Message 11214 - Posted: 22 Feb 2006, 14:01:15 UTC - in response to Message 11212.  
Last modified: 22 Feb 2006, 14:10:33 UTC

I have several machines with 256MB running Rosetta without any problems.....



It is certainly possible to run Rosetta with less than the recommended minimum system configuration. However, this can cause errors and Wu failures. One related issue would be page faults which can also cause Rosetta to fail.

EDIT: I just found this post from David Kim. Many of the errors you have are for this type of WU. The Batch has been cancelled. This should help your situation.



Moderator9
ROSETTA@home FAQ
Moderator Contact
ID: 11214 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
AMD_is_logical

Send message
Joined: 20 Dec 05
Posts: 299
Credit: 31,460,681
RAC: 0
Message 11226 - Posted: 22 Feb 2006, 23:44:40 UTC - in response to Message 11212.  

I have several machines with 256MB running Rosetta without any problems.....


256MB is fine for a dedicated cruncher. The 512MB recomendation is for the typical type of computer running rosetta, meaning one with some other purpose and where rosetta is supposed to quietly eat spare cycles without affecting performance. If the system has 512MB+ memory installed then the amount rosetta uses will be too small a fraction of that to have a significant impact on the system's responsiveness.
ID: 11226 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile [B@H] Ray
Avatar

Send message
Joined: 20 Sep 05
Posts: 118
Credit: 100,251
RAC: 0
Message 11236 - Posted: 23 Feb 2006, 10:30:58 UTC - in response to Message 11226.  

I have several machines with 256MB running Rosetta without any problems.....


256MB is fine for a dedicated cruncher. The 512MB recomendation is for the typical type of computer running rosetta, meaning one with some other purpose and where rosetta is supposed to quietly eat spare cycles without affecting performance. If the system has 512MB+ memory installed then the amount rosetta uses will be too small a fraction of that to have a significant impact on the system's responsiveness.


Really should go with 512 Meg or more for Rosetta. My HD started running hotter part the time Rosetta was running at first, increased ram from 512 to 1024 Megs and it colled off again. Run times also dropped by about 10%. 256 Megs will work but hit the swapp file to often and will burn out the HD.


Pizza@Home Rays Place Rays place Forums
ID: 11236 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Tom Lee

Send message
Joined: 10 Dec 05
Posts: 5
Credit: 129,728
RAC: 0
Message 11304 - Posted: 24 Feb 2006, 10:10:52 UTC - in response to Message 11168.  

I'm having more and more wu that end-up with a computation error status. I've had 15 hours worth in the last 24. I'm not sure what this means. My computer can't add, subtract, and divide, or it's not following program instructions? Do I have a rogue CPU? Could someone enlighten me - I can't find mention of it on the message boards. Thanks.



Check out this FAQ in the FAQs thread


The status specifically says "computation error". I can't find this in the FAQ. So what's happened when I get this message?



"Computation error" in the work display of BOINC is a catch all term used for a number of different problems.

If you look at the workunit in you stats page, you will find the "exception" errors in the errors reported for the WUs. What I am seeing is that with the exception of one of your errors, other systems have also failed to complete the same work units. This tells me that it may be a work unit issue. You really do not have that many errors. and the one that is unique to your system may also turn out to be work unit related but the other system that it was issued to has not reported back yet.

Based on what I saw on your stats page, I refered you to that particular FAQ. It is possible that there was some transient probelm on your system that has caused the one error that does not look like a WU issue.

Your system does seem marginal on memory, you might try increasing your swap time if you are running more than one project and see if that helps.

Thanks for the reply and the time you took to look-up my stats. It's nice to know that it's probably not my machine, but it is discouraging to see the computer time go down the drain. I'm also not sure what you mean by increasing the swap time. If you find time in your busy schedule, maybe you can fill me in. Thanks, Tom.
ID: 11304 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Whl.

Send message
Joined: 29 Dec 05
Posts: 203
Credit: 275,802
RAC: 0
Message 11306 - Posted: 24 Feb 2006, 10:37:07 UTC - in response to Message 11236.  

Ray Brown wrote:


Really should go with 512 Meg or more for Rosetta. My HD started running hotter part the time Rosetta was running at first, increased ram from 512 to 1024 Megs and it colled off again. Run times also dropped by about 10%. 256 Megs will work but hit the swapp file to often and will burn out the HD.


Putting the swap file on a second physical drive helps a bit also.

ID: 11306 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Astro
Avatar

Send message
Joined: 2 Oct 05
Posts: 987
Credit: 500,253
RAC: 0
Message 11310 - Posted: 24 Feb 2006, 11:53:29 UTC - in response to Message 11304.  

I'm also not sure what you mean by increasing the swap time. If you find time in your busy schedule, maybe you can fill me in. Thanks, Tom.


I'm not sure what he means but he may mean the "write to disk at most" setting. Below are the two possibly relevant setting in your "general preferences". (These are from my preference)

Processor usage
Switch between applications every
(recommended: 60 minutes) 180 minutes

Disk and memory usage
Write to disk at most every 600 seconds

ID: 11310 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Snake Doctor
Avatar

Send message
Joined: 17 Sep 05
Posts: 182
Credit: 6,401,938
RAC: 0
Message 11318 - Posted: 24 Feb 2006, 13:39:47 UTC - in response to Message 11304.  

I'm having more and more wu that end-up with a computation error status. I've had 15 hours worth in the last 24. I'm not sure what this means. My computer can't add, subtract, and divide, or it's not following program instructions? Do I have a rogue CPU? Could someone enlighten me - I can't find mention of it on the message boards. Thanks.



Check out this FAQ in the FAQs thread


The status specifically says "computation error". I can't find this in the FAQ. So what's happened when I get this message?



"Computation error" in the work display of BOINC is a catch all term used for a number of different problems.

If you look at the workunit in you stats page, you will find the "exception" errors in the errors reported for the WUs. What I am seeing is that with the exception of one of your errors, other systems have also failed to complete the same work units. This tells me that it may be a work unit issue. You really do not have that many errors. and the one that is unique to your system may also turn out to be work unit related but the other system that it was issued to has not reported back yet.

Based on what I saw on your stats page, I refered you to that particular FAQ. It is possible that there was some transient probelm on your system that has caused the one error that does not look like a WU issue.

Your system does seem marginal on memory, you might try increasing your swap time if you are running more than one project and see if that helps.

Thanks for the reply and the time you took to look-up my stats. It's nice to know that it's probably not my machine, but it is discouraging to see the computer time go down the drain. I'm also not sure what you mean by increasing the swap time. If you find time in your busy schedule, maybe you can fill me in. Thanks, Tom.


I think he means the time between application swaps. There is a setting for that in the prefs. I think the deafult is every 60 min, but you can set it to other amounts of time.

We Must look for intelligent life on other planets as,
it is becoming increasingly apparent we will not find any on our own.
ID: 11318 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Beezlebub
Avatar

Send message
Joined: 18 Oct 05
Posts: 40
Credit: 260,375
RAC: 0
Message 11336 - Posted: 24 Feb 2006, 16:11:24 UTC
Last modified: 24 Feb 2006, 16:13:18 UTC


Processor usage
Switch between applications every
(recommended: 60 minutes) 180 minutes

Disk and memory usage
Write to disk at most every 600 seconds


This will increase memory usage and make the program need more swap space which will mean slower wu times as writing to disk is vastly slower than RAM. The best solution for faster wu times is more memory, how much is up to you but 512 is the least I would use. If you still want more swap space go: start, settings, control panel, system, advanced, performance, advanced, virtual memory and adjust accordingly.

Keep in mind any increase in hard disk activity decreases lifetime so the more RAM you have the better.
e6600 quad @ 2.5ghz
2418 floating point
5227 integer

e6750 dual @ 3.71ghz
3598 floating point
7918 integer


ID: 11336 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
AXEL 70

Send message
Joined: 12 Dec 05
Posts: 1
Credit: 101,212
RAC: 0
Message 11341 - Posted: 24 Feb 2006, 17:08:17 UTC
Last modified: 24 Feb 2006, 17:23:04 UTC

excuse me for my english not very good (I'm Italian)...
the last WU's that I download (HBLR_1.0_... series)
give me all a computation error :

|rosetta@home|Unrecoverable error for result HBLR_1.0_1hz6_314_44_0 ( - exit code -164 (0xffffff5c))
|rosetta@home|Unrecoverable error for result HBLR_1.0_1hz6_314_40_0 ( - exit code -164 (0xffffff5c))
|rosetta@home|Unrecoverable error for result HBLR_1.0_1b72_314_76_0 ( - exit code -164 (0xffffff5c))

and that after 6:27, 1:45 and 0:50 hours ...

why? it's the first time that it happen
thank you for the answer (please answer simply)

p.s.: the client is 4.82
ID: 11341 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : Computation Error



©2024 University of Washington
https://www.bakerlab.org