Posts by old_user13614

1) Message boards : Number crunching : Announcement: Database residual problem - misallocated WUs (Message 13471) Posted 15 Jun 2005 by old_user13614 Post: My problem work unit has now completed. On the results page, you can get the graphs for phases 2 and 3, but not for phase 1, nor for the full run. I did back up the work unit folder last night just before completion, so if anything needs to be redone, it should be possible to pick up most of the way through the unit. All the messages in my client seem to indicate that the work unit uploaded successfully. Hope all is working smoother now.
2) Message boards : Number crunching : Announcement: Database residual problem - misallocated WUs (Message 13013) Posted 1 Jun 2005 by old_user13614 Post: Just to check a bit further - the misallocated WU I reported above (WU 565036) has now progressed into phase 2. On the results page, however, the graphs for temperature and precipitation in phase 1 are not available as links. In addition, on the results page, the "stderr out" does have a value indicating that the unit failed with exit code 5. My machine continues to crunch, and the trickles keep piling up. My suspicion is that the important data is all here on my machine - but the bogus exit code (perhaps submitted by another user on this unit - perhaps just some other wierd glitch) is preventing the graphs from being available. I am happy to do whatever is most appropriate (abandon the model / stop and backup the data until the problem can be addressed / just keep crunching / ...) and realize that no one may really have the best answer at this point. If someone wants to contact me directly, my email is djd at isd dot net (slightly disguised from the spammers spiders.) Thanks,
3) Message boards : Number crunching : Announcement: Database residual problem - misallocated WUs (Message 12792) Posted 23 May 2005 by old_user13614 Post: I think I have one of these: WU: 565036 Work Unit name: 2u1q_300153589_1 My Host ID: 22439 Host ID identified on results page: 48533 Completed 6 trickles - computer is showing that it is working on TS 75001 when I checked seconds ago, so a 7th trickle will come through shortly. If the crunching my computer is doing will actually be valuable - that is more important than getting the credit - and I will just let it keep going. Thanks for all your work. Lornix
4) Message boards : Number crunching : CPU Performance down under 4.19 for Windows? (Message 9852) Posted 23 Feb 2005 by old_user13614 Post: My sec/TS value had been pretty consistently in the 2.7-2.8 range on my Windows machine running either 4.05 or 4.13 However, since upgrading to 4.19 in late January - the value has climbed up to the 2.95 - 3.05 range. Since I upgraded shortly after completing a run, it could be that I just have a new model with a set of parameters that require a bit more computation. Has anyone else seen any change in their speed under 4.19? If so, do we know whether that is a "good" hit (e.g. an increase in computation caused by a more reliable error checking procedure or some such), or a "bad" hit (e.g. an increase caused because the BOINC developers don't do as much work optimizing a Windows version for speed.) For what is worth, I don't seem to notice any comparable performance hit since upgrading my (noticeably slower) Linux system to 4.19 - so if there is anything going on, it may only affect Windows machines.
5) Message boards : Cafe CPDN : Question for the climatologists (Message 4085) Posted 14 Sep 2004 by old_user13614 Post: As one with just about exactly the right \"little knowledge\" of science to be dangerous - I have been wondering about the following. According to the story I heard about the \"butterfly effect\", it was discovered when a climatologist had to interrupt a long climate model, so took a printout of values from one point in the middle of a run, and simply input those values to start another run when he could come back. He was shocked to see a different set of results come out, and upon further investigation, he discovered that the different pattern arose because the printouts were only displaying 8 digits while the internal memory of the computer was carrying the values to greater precision, and thus the values were different by very small amounts. I understand this to imply that very small differences in values like wind speed can have some impact on weather patterns. The second consideration is simply thermodynamic - energy can not be created nor destroyed (as long as we understand mass as carrying the potential energy identified in Einsteinian relativity theory). As a very specific example, this would imply that a wind energy farm, that uses wind to generate electricity must be extracting at least some of the energy of the wind, presumably by reducing its velocity. The question that I have is, given reasonable estimates of the efficiency of wind turbines, how much energy can we extract from wind before that would start to have effects on our weather. Has anyone ever done even simplistic examinations of this kind of topic? If so, what kinds of results have they obtained? Thanks for any answers you might have.
6) Message boards : Number crunching : CPDN Classic Stats Brought Over To BOINC (Message 3981) Posted 12 Sep 2004 by old_user13614 Post: I have set the classic CPDN user name for my "Home" machine to be the user name I used to use (and signed up for BOINC with the same user name) - but I see no sign of my classic credits. I can imagine a few different reasons for this - but one of the most obvious would be that the update is no longer happening, or is happening far less frequently, so before I try lots of different alternatives, I want to see if you are still doing that update procedure. The second obvious category of problem is that I had used two different user names - it was set as my e-mail address by default when I signed up, and it was only after a month or so that I switched over to my current user name. Since the form says something like "original CPDN classic user name" it is a bit unclear whether the critical thing is the most recent user name, or the first name ever used. Since I was credited with 7 complete runs under CPDN classic, I would kind of like to get that credit back, but don't want to push too hard. Feel free to contact me directly via e-mail if that is more appropriate Don Dresser user name: lornix e-mail: dee jay dee at eye ess dee dot en eee tee