Posts by dajashby

41) Questions and Answers : Unix/Linux : Model crashed with 1 trickle to go (Message 14315) Posted 12 Jul 2005 by dajashby Post: When I posted, the mark up codes got left out: "4.13" "process got signal 11" "" "3" "11" "" "" I hope this works...
42) Questions and Answers : Unix/Linux : Model crashed with 1 trickle to go (Message 14314) Posted 12 Jul 2005 by dajashby Post: The result ended with this stderr-out. 4.13 process got signal 11 3 11 I believe this occured while the project was down on the 10th or 11th of July. I don\'t know how to tell whether the model actually completed. Is there any way we can rescue this? The computer concerned is currently running a newly downloaded model. Running boinc 4.13 (yes, I know, I should upgrade ...) Host no is 73339 Model name is 27cq_300123887_0 (Result 720418)
43) Message boards : Number crunching : Anyone else experience decrease in team credits? (Message 14171) Posted 6 Jul 2005 by dajashby Post: > Supporting the existing work is important, but so is being able to move on to > other important and exciting new experiments. There is so much that is > important that it sometime seems that maintenance has to reach critical levels > of importance to get attention. This is a modern disease - in the long run maintenance is always cheaper than repair, but people never think in the long term, which is why global warming is happening in the first place!
44) Message boards : Number crunching : Announcement: Database residual problem - misallocated WUs (Message 12994) Posted 31 May 2005 by dajashby Post: I have a similar problem with my host 71544, which seems to have hijacked result 865168 from host 92060. It has contributed 9 trickles since 26th May but received no credit for them, and the result is not listed in its result list. Should I abort the result by resetting the project?
45) Questions and Answers : Unix/Linux : Ran out of disk space... (Message 5840) Posted 1 Nov 2004 by dajashby Post: > Ouch! > > Sounds like a corrupt file has rewound your model for you. If you have a > fairly recent backup of your BOINC directory you could restore that so that > you have to redo less of the model. > > From a science point of view you should continue with the model, but if you > can't live without getting credits for the work your system is doing you could > do a project reset to force an upload of the restarted job and get a new one. > If you do decide to that the workunit will be rescheduled for someone else to > run. > <br><a href="http://www.teampicard.net"><img> src="http://www.teampicard.net/templates/fisubice/images/phpbb2_logo.jpg"></a><a> href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/team_display.php?teamid=3">Join > us here</a> > Thanks - I'm afraid I went for the project reset option before getting your reply. Up until now I haven't been backing up my boinc files, but I'll clearly have to consider doing this. I've offloaded a heap of music files from the linux box, and now have plenty of space on it. (Moonlight - the Windows machine - has a 120 GB hard drive but plato only has a 40.)
46) Questions and Answers : Unix/Linux : Ran out of disk space... (Message 5815) Posted 31 Oct 2004 by dajashby Post: My model was into phase 3 when it attempted a trickle. Unfortunately I had succeeded in filling up the user area diskspace. The request to the server failed, with an empty file message, and the model hung. When I restarted it began the model from scratch (i.e Phase 1), and has since refused to trickle (up to Timestep 17683 as I send this.) What can I do? (Apart from taking more care in future?)
47) Questions and Answers : Windows : Trickles not being uploaded as of Sunday 10/10/04 (Message 5213) Posted 11 Oct 2004 by dajashby Post: I have a Windows PC and a Linux PC that are both having the same problem - they both tricked within an hour of each other before 3:00 UTC on the 10th, and have not done so since. I would normally expect both to trickle about twice a day. Communication to the server seems to be okay on both.
48) Message boards : Number crunching : Leaderboard gripe (Message 5203) Posted 10 Oct 2004 by dajashby Post: > Maybe this helps: > > <a> href="http://www.boincstats.com/stats/host_graph.php?pr=cp&id=11065">http://www.boincstats.com/stats/host_graph.php?pr=cp&id=11065</a> > > <a href="http://www.boincstats.com/"><img> src="http://www.boincstats.com/stats/banner.php?cpid=a4c31845a9413499815907bae59136ac" > border="0"></a> > Thanks very much for that. Almost more statistics than I will ever need. I have just one criticism - the Hosts list doesn't include the HostID. I thought that was a little odd.
49) Message boards : Number crunching : Computer performance when running models (Message 5175) Posted 10 Oct 2004 by dajashby Post: > In your case I would make BOINC run only 1 model at a time. It will complete > that one faster than when running 2 and you won't have to think about starting > and stopping stuff. > > When you alter you settings to 1-model-only-mode the client will wait untill > next trickle (or untill you press "update prefs") and will then pause model 2 > untill model 1 is finished - so you won't loose any work. > Thanks for the info.
50) Message boards : Number crunching : Computer performance when running models (Message 5171) Posted 10 Oct 2004 by dajashby Post: > > BOINC recognizes HT processors as 2 processors and will run one model per > CPU. > > The old version would only run 1. This seems to be what you are seeing. > If you > > only want to run one at a time you can set maximum number of CPUs to 1. > > <br>John Keck -- BOINCing since 2002/12/08 -- <a> > href="http://www.boinc.dk/index.php?page=user_statistics&project=cpdn&userid=191"><img> > border="0" height="80" src="http://191.cpdn.sig.boinc.dk?188"></a> > > > > > I wondered if that might be the case. If I now go in and change that setting, > what will happen to the second model? Will it continue to run until it's > finished, or will it sit there until the first one is finished and then > continue, or is it okay to cancel one of the models - and if so, how do I do > that? > > On reflection, there isn't a problem with running two models on the machine when I am not using it. I currently have the preferences set to suspend the model when I'm doing anything else on the computer, but running one model at such times is not really a problem. Can I manually suspend one model, and not the other, I wonder? The only problem there is remembering to unsuspend it when I stop using the computer for anything else.
51) Message boards : Number crunching : Computer performance when running models (Message 5169) Posted 10 Oct 2004 by dajashby Post: > BOINC recognizes HT processors as 2 processors and will run one model per CPU. > The old version would only run 1. This seems to be what you are seeing. If you > only want to run one at a time you can set maximum number of CPUs to 1. > <br>John Keck -- BOINCing since 2002/12/08 -- <a> href="http://www.boinc.dk/index.php?page=user_statistics&project=cpdn&userid=191"><img> border="0" height="80" src="http://191.cpdn.sig.boinc.dk?188"></a> > > I wondered if that might be the case. If I now go in and change that setting, what will happen to the second model? Will it continue to run until it's finished, or will it sit there until the first one is finished and then continue, or is it okay to cancel one of the models - and if so, how do I do that?
52) Message boards : Number crunching : Computer performance when running models (Message 5167) Posted 9 Oct 2004 by dajashby Post: I\'m convinced that the BOINC client is comsuming far more of my cpu cycles than the Classic CPDN did. Generally speaking CPDN usually took up about 40% of my CPU, BOINC is often at close to 100. The problem may be that I appear to be running two models. Is this normal, or did I stuff up when I was setting up the process? If it isn\'t normal to run two at a time, how can I reduce the load without causing problems? I\'m running XP SP2 on a 2.8 Ghz Pentium 4 machine, and I\'ve currently got preferences set to suspend activity when I\'m using the machine.
53) Message boards : Number crunching : Leaderboard gripe (Message 5166) Posted 9 Oct 2004 by dajashby Post: My main desktop computer is at the moment listed at No 353 on the Top Computer\'s leaderboard. I realise just how trivial this is, but I\'m a big believer in software useability. Could we have the Next 20, Last 20 commands at the top of the page, as well as the bottom? All the wheel scrolling I need to do to get down to them in order to check where I am is wearing out my mouse! Alternatively, would it be possible to put the leaderboard position on the account summary screens?
54) Questions and Answers : Unix/Linux : Permission denied when running linux client (Message 5143) Posted 8 Oct 2004 by dajashby Post: > Is the "execute" permission set? Try: chmod +x [boinc_filename] > > Also try "ls -l" to check the permissions. It should look something like: > -rwxr-xr-x 1 root boinc 9458934 Sep 17 17:33 > boinc_4.09_i686-pc-linux-gnu > > Thanks for that solution. *Blush* When I looked at the file's permissions in the KDE file manager I took the Execute checkbox to be a selected toggle button. Just shows you that GUIs aren't the answer to everything.
55) Questions and Answers : Unix/Linux : Permission denied when running linux client (Message 5129) Posted 7 Oct 2004 by dajashby Post: I downloaded the boinc linux client, and ran Ark to extract the files. I only get one executable file, and when I try running it at the console I just get a Permission denied message. I downloaded the file as my standard user, and I defintely own the file. I\'m running gentoo linux with a 2.6 kernel. The computer is a Pentium 4. The client is 4.09

Previous 20