Message boards : Number crunching : problem with boinc 4.45 or climate prediction?
Message board moderation
Author | Message |
---|---|
Send message Joined: 19 Jun 05 Posts: 15 Credit: 54,022 RAC: 0 |
During pre-boinc climate prediction I finally quit after two weeks with no trickle and no answer about what the problem was. (I don\'t mind risking 10 or even 100 hours, but not three months. It also made me nervous the talk of searching for files in the hard drive to manually upload.) I am now giving boinc climate another chance. After a few hours the first boinc climate crashed. I downloaded another boinc climate. BOINC now won\'t download any other projects and I am only running climate. Would it help to go to 4.19? Wait for a new boinc version? Wait for a fix from climate? I do not mind if I am not updated every day with a trickle on this project, but I should have something within a week or so. If I can\'t resolve this within 2 weeks or so I will give up all together on climate. Months of computing time is just too valuable to risk. <img> |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
It's impossible to guarantee anything on this project. Lots of people, with lots of different harware/sofware, have repeated failues for no obvious reason. And lots of people, with lots of different harware/sofware, have NO problems. Then there are some who start out with a problem, that someone here manages to solve and get them going. Just to cheer you up, a few people recently have posted that their model has failed on the final trickle. The refusal to download, if accompanied by the message "reached daily limit", is to stop fast occuring, continual failures from quickly draining the data set pool. BOINC 4.19 seems to be the most stable version for a range of problems. E.G. for people with an ATI graphics card. I'm running 4.25 OK. Versions 4.4n onwards are a bit temperamental. There's no point in waiting for a fix from CP, as there are too many variables, and the main culprit is usually BOINC combined with user's hardware. And there is currently a problem with the CP servers. Trickles, credits, and xlm stats files aren't working, for unknown reasons. Your hardware/software seems to be suitable; the only unknown is which graphics card you have. Best I can advise is, give 4.19 a go and see what happens. If you have more problems, post again. |
Send message Joined: 16 Feb 05 Posts: 2 Credit: 166,200 RAC: 0 |
> It's impossible to guarantee anything on this project. Lots of people, with > lots of different harware/sofware, have repeated failues for no obvious > reason. And lots of people, with lots of different harware/sofware, have NO > problems. > Then there are some who start out with a problem, that someone here manages to > solve and get them going. Just to cheer you up, a few people recently have > posted that their model has failed on the final trickle. > > The refusal to download, if accompanied by the message "reached daily limit", > is to stop fast occuring, continual failures from quickly draining the data > set pool. > > BOINC 4.19 seems to be the most stable version for a range of problems. E.G. > for people with an ATI graphics card. I'm running 4.25 OK. Versions 4.4n > onwards are a bit temperamental. > > There's no point in waiting for a fix from CP, as there are too many > variables, and the main culprit is usually BOINC combined with user's > hardware. > And there is currently a problem with the CP servers. Trickles, credits, and > xlm stats files aren't working, for unknown reasons. > > Your hardware/software seems to be suitable; the only unknown is which > graphics card you have. > > Best I can advise is, give 4.19 a go and see what happens. If you have more > problems, post again. > > > |
Send message Joined: 16 Feb 05 Posts: 2 Credit: 166,200 RAC: 0 |
I went from 4.19 to 4.25 to 4.44, and to reccommended 4.45. The last one didn't behave correctly particularly in non-English environment so I returned to 4.44 that runs satisfactorily last 10 days. |
Send message Joined: 14 Mar 05 Posts: 58 Credit: 55,577 RAC: 0 |
> During pre-boinc climate prediction I finally quit after two weeks with no > trickle and no answer about what the problem was. (I don't mind risking 10 or > even 100 hours, but not three months. It also made me nervous the talk of > searching for files in the hard drive to manually upload.) I am now giving > boinc climate another chance. After a few hours the first boinc climate > crashed. I downloaded another boinc climate. BOINC now won't download any > other projects and I am only running climate. Would it help to go to 4.19? > Wait for a new boinc version? Wait for a fix from climate? I do not mind if I > am not updated every day with a trickle on this project, but I should have > something within a week or so. If I can't resolve this within 2 weeks or so I > will give up all together on climate. Months of computing time is just too > valuable to risk. > <img> >I feel the same way also.360 hours of computer time and one Phase 3 also. Cannot afford the chance of losing this Data! Sincerely Doug Worrall> |
©2024 cpdn.org