| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: The Clean Energy Project - Phase 2 Forum
Thread: widoze XP crash .... |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 7
|
|
| Author |
|
|
JoeCyber
Cruncher Joined: Apr 20, 2007 Post Count: 40 Status: Offline Project Badges: 
|
i found something strange ... after 138 days of works on this project.
----------------------------------------on my smaller computers (3) i still run windoze xp sp3 and WCG 6.10.58 and until recently i wasn't really using them on this project, i mostly work this project on bigger computer on vista. but few weeks ago i try switching all computers on this project ... i found systematically my windoze xp crashed after 2 or 3 days of computing without a reboot... other OS (like vista) run without problem ... my xp computer can run all week long computing other projects( cancer mostly) ... there's something really strange here, maybe some programming errors (pointer leak, file pointer leaks etc ...)who has no effect on newer OS but bring my 3 xp machines down ... is it just me ? anyway i'm back on crunching cancer for a while :-) [Edit 1 times, last edit by JoeCyber at May 13, 2011 7:26:03 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Where to begin finding out why the 3 smaller crash after running few days on CEP2? Lets start with finding the stdoutdae.txt file in the BOINC data directory and posting a copy of the log entries, from before it crashes, up to inclusive when the tasks are restarted. Do they restart or are they then reported with error and if with error, can you plz visit the Result Status page and click such an error link. Then post a copy of the Result log, so we can see if there's anything pointing.
----------------------------------------ttyl --//-- [Edit 1 times, last edit by Former Member at May 13, 2011 7:59:19 AM] |
||
|
|
JoeCyber
Cruncher Joined: Apr 20, 2007 Post Count: 40 Status: Offline Project Badges:
|
stdoutdae.txt seems to be an one month rolling log file, my units were more that one month old ...
when my computers crashed i had to hard reset them, so that means i had to scandisk them each time ... i don't like to have hole on my files system ... all the units before crashing were thrashed with error... if i remember, unable to open a file... i don't know if i want to recrashed one my comp again ... i have to physically move it to one point on my LAN where a screen is available ... (im running them without moniter and keyboard). if i have time i will and report back... but hey if im they only one who report this ... i might just continue crunching cancer units without problem ... :o) |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
stdoutdae.txt seems to be an one month rolling log file, my units were more that one month old ... The deadline for most WCG tasks is 10 days, so having/computing tasks older than 1 month is an effort in futility. Copies will long have been re-distributed to other crunchers to complete. On starting a computer with such old tasks there are actually red-line warnings in the message log to say they've long expired. Long as the tasks are still on the server, the client might even receive an instruction to cancel the task, but only if not already started.when my computers crashed i had to hard reset them, so that means i had to scandisk them each time ... i don't like to have hole on my files system ... all the units before crashing were thrashed with error... if i remember, unable to open a file... i don't know if i want to recrashed one my comp again ... i have to physically move it to one point on my LAN where a screen is available ... (im running them without moniter and keyboard). if i have time i will and report back... but hey if im they only one who report this ... i might just continue crunching cancer units without problem ... :o) With such old tasks it's improbably to find the Result Error log on the website, but it on the Result Status page where you find them, clicking the Status column links. The stdoutdae.txt file holds X bytes of messages, 1 or 2 MB, can be configured to hold more. If the file is full, a copy is created to stdoutdae.old and the stdoudae.txt is emptied. There are [free] pieces of software to remote control other computers long as they're on a network. Your machines might be running hot after a few days. Disconnect power and open case to remove all dust carefully. Compressed air or a soft brush. CEP2 is amongst the heaviest on the grid, so maybe HCC might be the better option, if it's not dust. Let us know. --//-- |
||
|
|
JoeCyber
Cruncher Joined: Apr 20, 2007 Post Count: 40 Status: Offline Project Badges:
|
stdoutdae.txt seems to be an one month rolling log file, my units were more that one month old ... The deadline for most WCG tasks is 10 days, so having/computing tasks older than 1 month is an effort in futility. Copies will long have been re-distributed to other crunchers to complete. On starting a computer with such old tasks there are actually red-line warnings in the message log to say they've long expired. Long as the tasks are still on the server, the client might even receive an instruction to cancel the task, but only if not already started.when my computers crashed i had to hard reset them, so that means i had to scandisk them each time ... i don't like to have hole on my files system ... all the units before crashing were thrashed with error... if i remember, unable to open a file... i don't know if i want to recrashed one my comp again ... i have to physically move it to one point on my LAN where a screen is available ... (im running them without moniter and keyboard). if i have time i will and report back... but hey if im they only one who report this ... i might just continue crunching cancer units without problem ... :o) With such old tasks it's improbably to find the Result Error log on the website, but it on the Result Status page where you find them, clicking the Status column links. The stdoutdae.txt file holds X bytes of messages, 1 or 2 MB, can be configured to hold more. If the file is full, a copy is created to stdoutdae.old and the stdoudae.txt is emptied. There are [free] pieces of software to remote control other computers long as they're on a network. Your machines might be running hot after a few days. Disconnect power and open case to remove all dust carefully. Compressed air or a soft brush. CEP2 is amongst the heaviest on the grid, so maybe HCC might be the better option, if it's not dust. Let us know. --//-- eeee .... my units were probably 1or2 days old when i compute them.... what i said was old is the log into the stdoudae.txt , again it seems to be a one month log and i made the swith back to other project more maybe 1.5 month ago so log of my unit are not there... yes there are free software for controlling my computer, i user ultravnc, but when my computer crash i should said it jammed XP but not completely , no blue screen, more like when your program allocate many many resources and never free them (example opening zillion of files without closing them) so xp hang itself after a while ... of course at those time, ultravnc got trash too, so i had to hard reset the computer ... so even after i moved one machine to a place with a screen and keyboard it was still hanging the same way and i found that it was hanging with keyboard frozen so i add to hard reset them ... nope my machines are fine they don't overheat, i know how to build computer :o), and i should also say that one of my xp was also plug on UPS so it is not weird electric glitch ... we will see maybe ... later then  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I run Cep on my xp, sp3 with a single core cpu 1 GB Ram with no problems, freezes/crashes etc. Last reboot 57 days ago, just saying.....
|
||
|
|
JoeCyber
Cruncher Joined: Apr 20, 2007 Post Count: 40 Status: Offline Project Badges:
|
I run Cep on my xp, sp3 with a single core cpu 1 GB Ram with no problems, freezes/crashes etc. Last reboot 57 days ago, just saying..... i though about few other potentiel problem ... i might investigate in deep some days ... thanks :-) |
||
|
|
|