Posts by jujube

21) Message boards : News : Very long jobs (Message 24710) Posted 24 Aug 2012 by jujube Post: Du calme as we say in French. I am hurt by your accusation. I have always insisted that we run on any "old" machine, even a Pentium 3. The experiment with 10 million turns has shown some deadline/scheduling problems. These will be addressed before we proceed. The few hundred jobs are necessary to complete a couple of studies. I have a few hundred thousand other jobs available. Watch this space. Eric. Tasks that are much longer than the tasks a project usually issues cause problems on some hosts. Also, it's entirely possible that your longest tasks won't finish before deadline on a Pentium 3 even if it crunches 24/7. Other projects have run into exactly the same problem and rather than keep all tasks short they have configured their server to offer 2 different types of tasks: long tasks and short tasks. They create an option in user website prefs to allow crunchers to choose both types or just one type depending on their hardware and expectations. I have no idea how much work it would take to implement such an option (perhaps server code already has that functionality?) but you/Igor might give it some thought. GPUgrid offers a long vs. short option, maybe they can give you/Igor some ideas on how to implement it.
22) Questions and Answers : Sixtrack : no work (Message 23939) Posted 19 May 2012 by jujube Post: Because there is no work. This project has work once in a while but not very often.
23) Message boards : Number crunching : Virtual machine set up (Message 23938) Posted 17 May 2012 by jujube Post: Chris, This project was started for the purpose of simulating the paths of particles in the LHC for the purpose of tuning the magnets that focus the beam. That job has been completed for now but they plan to upgrade the LHC and use different particles than they are using now. Both of those factors will require tuning the magnets differently so there is a good chance there will be more work here eventually but nobody knows for sure when that campaign will start. Nor has there been any promise that the job will definitely be given to Sixtrack. From what I've heard in the mainstream media, I doubt it will start in 2012 but ya never know. The only sure thing is that the people who know aren't talking.
24) Questions and Answers : Wish list : I want to see more involvement from CERN. (Message 23903) Posted 28 Feb 2012 by jujube Post: They don't have miles of data here so there will never be continuous work.
25) Message boards : Number crunching : Problems with uploading (Message 23900) Posted 24 Feb 2012 by jujube Post: The server that receives the uploads may be offline or there may be a DNS problem. The situation will probably correct itself in a few hours. In the meantime there is noting you can do but be patient.
26) Questions and Answers : Wish list : VM Startup (Message 23897) Posted 23 Feb 2012 by jujube Post: This project doesn't require a VM. You must be thinking of this project's sister project, Test4Theory@home, which is a CERN sponsored project like this one. Test4Theory has a constant supply of work whereas this project has work intermittently.
27) Message boards : Number crunching : Server status page shows 0 tasks ready but work is issued (Message 23894) Posted 19 Feb 2012 by jujube Post: This could be what I'm noticing then, but in reverse. The server status page is showing over 1,000 tasks, but the client isn't getting. I was wondering if it might have been just tasks for another platform available or something, maybe not... To understand why that happens you need to understand a fact that is repeated over and over and over in these threads and the threads from most other BOINC projects. I'm surprised you haven't heard this fact before since you've been with this project since 2005. OK, ready? Here it is... The "tasks ready to send" number you see on the server status page is not accurate. Why? Because the server does not count the number of tasks waiting whenever you click the server status link. The server counts the tasks ready to send every 15 minutes or so and remembers that number and displays that number to everybody who clicks the server status link before the nexst time the tasks ready to send are counted. Read that again. Make sure you understand what that means. Now here's an example to illustrate.... Let's say the server counts the tasks every 15 minutes and let's say the last count was at 10:00 AM and there were 1500 tasks. The next count won't be until 10:15 AM. Now suppose you click the server status link at 10:14 AM. How many ready to send tasks will you see? You'll see 1,500 because that's how many the server found when it counted at 10:00 AM. So here's the million dollar question..... If its says 1500 tasks ready to send then why can't you get any? The answer is simple. The answer is because there are no tasks ready to send. But, but, but doesn't the server say 1,500? Yes but that was 14 minutes ago at 10:00 AM when the server last counted. Between 10:00 AM and 10:14 AM several thousand computers requested work and were given some. In fact they gobbled up ALLnthe tasks before your computer requested tasks. So when the server counts the available tasks again at 10:15 it will say 0 tasks ready to send. Still don't understand? Then read it again because it's really quite simple and I've explained it top to bottom, nothing left out, as simple an explanation of this mystery as you will get anywhere. Now the 10 million dollar question.... If there were no tasks ready to send then why was Tom able to get some? The answer is that some of the 1000 sent crashed and needed to be resent. They were gobbled up so quickly by the many thousands of computers requesting work they were not counted by the server.
28) Questions and Answers : Preferences : Detailed Settings of Preferences (Message 23884) Posted 10 Feb 2012 by jujube Post: The settings are a bit confusing. a and b refer to the the number of cores/processors you allow BOINC to use. If you have a 4 core i7 with hyper threading enabled BOINC will see 8 cores. If you set a to 8 BOINC will use all 8. If you set a to 5 BOINC will use 5. If you set b to 100% BOINC will use 8. If you set b to 25% BOINC will use 2. I use the local settings rather than the website settings (click Tools -> Computing Preferernces -> CPU Usage) and those settings do not include item a, just b. I don't know why the website has a. If you use the local settings then they will override the website settings. Item c determines how many of your spare CPU cycles BOINC will use. It's a throttle and it's intended use is to reduce CPU temperatures. At 40%, BOINC will suspend computing for all your projects for 60% of the time and allow computing for 40% of the time. It's a very coarse throttle which operates at the seconds level as opposed to the milliseconds level. At 40%, for example, it will run the tasks for 4 seconds then suspend the tasks for 6 seconds. Regarding the a and b settings it's actually not correct to say "BOINC will use x cores" but it's become part of the BOINC culture to use that expression. What "will use x cores" actually means is that BOINC will run x single-threaded tasks at a time. How many cores the x tasks actually use is determined by the OS not BOINC. For example, if you have a 4 core CPU and allow BOINC to use only 1 core BOINC will run only 1 task but you may find that task running on 3 or 4 cores not just 1 core. BTW, the program is called BOINC, not BIONIC. And you will find this project, Sixtrack, does not have a continuous supply of tasks. There are long dry spells where there is no work available. If you are interested in particle physics and would like to help CERN with some LHC related work then consider joining Sixtrack's "sister" project Test4Theory@home which has a continuous supply of tasks. Be sure to follow the first few links on the home page to get to the directions for installing VirtualBox. Test4Theory will not run without VirtualBox.
29) Message boards : Number crunching : LHC@home Won't Work (Message 23877) Posted 2 Feb 2012 by jujube Post: What makes you think it isn't working? What, exactly, is happening or not happening that leads you to believe it's not working? Do you look in the Event Log after you click Update? That's the first place you should look whenever there is a problem. Don't bother reseting the project or reinstalling BOINC. That won't fix the problem you are having.
30) Message boards : Number crunching : Why isn't there much work? (Message 23876) Posted 2 Feb 2012 by jujube Post: The only thing to do is wait until the project has work available. They issue a batch of tasks and wait until they are all completed and verified. Before they can issue a new batch they need time to analyze the results from the previous batch. That can take weeks or even months.
31) Message boards : Number crunching : Need some work please (Message 23873) Posted 1 Feb 2012 by jujube Post: There is lots of work going on at the LHC. If you want more work related to the LHC then consider joining Test4Theory@home which is this project's sister project, both sponsored by CERN. Follow the links on the home page carefully because you must install VirtualBox in order to run T4T (Test4Theory).
32) Message boards : Number crunching : Why isn't there much work? (Message 23872) Posted 1 Feb 2012 by jujube Post: The project is not defunct. See the thread titled "No more work?".
33) Questions and Answers : Getting started : ??????????? (Message 23868) Posted 28 Jan 2012 by jujube Post: There is no Attach to a Program Page. Maybe you mean the Attach to a Project Wizard. Anyway you do have a computer attached to this project because it's in the list of your computers.
34) Questions and Answers : Windows : Error: "Runtime Error! ...request to terminate it in an unusual way..." (Message 23867) Posted 28 Jan 2012 by jujube Post: First thing you need to realize is that the server status page is a cached paged which is updated every 15 minutes or so. That means the server does NOT count the number of tasks ready to send when you click the server status link. The server counts the ready to send tasks when it feels like it (every 15 minutes or so) so the reported number of tasks ready to send is the number that were ready to send a few minutes ago which is probably not the number ready to send when you click the link. That's why you can see "100 ready to send" then click Update in BOINC manager and get "Project has no jobs available" in the log. Another thing you need to realize is that this project rarely has any work. That means if some project's science application is failing the odds are it's not this project's science app. If you want to know which app is failing wait for that useless error message then look in BOINC's log for a message indicating a task failed with no finished file. I have a hunch it's not this project's app because your list of tasks indicates you haven't run a task for a while. Another thing you can check is Windows' system event logs when you see that message popup. Sorry, I have no idea where those logs are because I don't run Windows.
35) Message boards : Number crunching : Faulty Computers or Modified BOINC ?? Huge Credits (Message 23817) Posted 8 Jan 2012 by jujube Post: Stelf, The problem with your RAC and total credits is not a result of an error when migrating you to Sixtrack's new location. The problem is caused by the huge CPU times for 4 results you recently submitted. Here is a list which contains the 4 results in question, they are numbered 2159845, 2159679, 2151335 and 2139853 in the lefthand column. Those 4 tasks probably didn't even exist when you were migrated but even if they did the CPU time numbers assigned to the tasks when they were created would have been overwritten (replaced) by the CPU times reported by your computer after it finished the tasks. I am not on a witch hunt and I am not a credit cop. I couldn't care less if you and every other Sixtrack volunteer cheats. I don't want to have you banished from the project. My purpose is to determine the facts and present them as clearly as possible in order to enlighten anybody who wishes to be enlightened about credits and credit cheating. As I said, the CPU times recorded for the 4 results came from your computer. How do I know they are not a glitch or due to a bug? Let's define glitch and bug first. A glitch is a one time event. It happens once due to a power spike, a gamma ray penetrating your CPU/HD and altering a bit/byte and other freak accidents. A bug is an error in the logic in the software that works with the data. Let's examine glitches and bugs as they pertain to your 4 results and see if we can rule out a glitch or bug as the cause of the 4 huge CPU times. A glitch happens in an instant and then it's gone. It's totally random and unpredictable. Can a glitch happen again later? Of course it can, it can happen any number of times BUT for a glitch to happen 4 times and to affect 4 results from Sixtrack but not results from other projects and affect ONLY the CPU times and only the CPU times for YOUR results (in the case of a glitch on the server), well, that is so statistically unlikely we can say it's impossible. Glitches are not that selective. Could it be a bug? If it were a bug then it would be affecting many people's results because all of our results, not just yours, are processed by the same purportedly buggy software. Of the many thousands of results from the most recent batch of tasks, this "bug" has affected only YOUR results and ignored everybody else's results. Not only that the "bug" assigned exactly the same CPU time to 2 different results and very similar CPU times to the other 2 results. A bug would tend to write exactly the same CPU time for all your results or it would write totally random numbers but it's obvious that neither of those is the case. If you think so then why does it not write negative numbers and why does it not write exceedingly huge numbers? For me, the only acceptable explanation for the 4 huge CPU times is that you altered the CPU times on your computer before they were reported to the server. It's easy to do and I've done it myself so I know it can be done. Again, I don't want to see you banished from Sixtrack and I don't even care if the project admins correct the 4 errant results. I hope you will reverse your decision to not accept Sixtrack tasks. If you don't there is no way to investigate the issue further and no way to prove it's a glitch/bug (though IMHO it's already proven it isn't).
36) Message boards : Number crunching : Faulty Computers or Modified BOINC ?? Huge Credits (Message 23809) Posted 7 Jan 2012 by jujube Post: It seems to me he could be running a 32-bit machine and if he sets the clock back to before the task started it will essentially be the unix time clock problem. If you go back 1 second you're logically going forward 2^32 seconds. Think Y2K only for unix time. If they were running 64-bit? Perhaps their machine is 64-bit but it's running the 32-bit app. The science app doesn't keep track of the time, BOINC does, so it makes no difference which science app he ran. In the case of the second cheater Kai links to, the one with 4 high credit tasks, note that only the CPU time is affected. If it were due to the clock being set back then surely the elapsed time would have been affected too, I think. Also, note the sent and returned times on the 4 tasks. 2 tasks were sent together, crunched and returned at the same time. Then another 2 were sent, crunched and returned. If the high CPU times are due to the clock being turned back after the tasks started then the clock must have been turned back twice, once while the first pair of tasks were crunching and again when the second pair were crunching. I could believe one clock adjustment but 2 in a row is not likely. However, the biggest reason I don't believe the clock turn back theory is that it doesn't seem to affect a task's CPU time or it's elapsed time. I tried it and it has no effect but maybe I'm doing it wrong or maybe it's just my system. Can anybody out there make it work?
37) Message boards : Number crunching : Faulty Computers or Modified BOINC ?? Huge Credits (Message 23805) Posted 6 Jan 2012 by jujube Post: The cheats for the tasks Kai has pointed out were accomplished by spoofing the elapsed and CPU times reported by BOINC client. About a week ago David Anderson released a server update or patch that prevents this type of cheat. If the Sixtrack admins can incorporate that patch/update into the current Sixtrack server setup then the "elapsed time and CPU time spoof" cheat will no longer work at Sixtrack.
38) Questions and Answers : Windows : Project still alive? (Message 23800) Posted 5 Jan 2012 by jujube Post: This project is still alive. There has been lots of work since the move back to CERN but it has not been continuous work. There are still long periods of time between batches of work units, however you should have received some work by now. After the move back to CERN there was some confusion/changes with the URL so perhaps you are not attached to the correct URL. The correct URL is given on the home page. The URL in the Add A Project Wizard is also the correct URL. Please note there is no work available at the moment and there might not be for a few more weeks. This project does not require VirtualBox. There is another project from CERN called Test4Theory@home and it is related to the LHC too. That is the project which requires VirtualBox and there is unlimited work from that project, it never runs out.
39) Message boards : Number crunching : Do we have a new URL? (Message 23798) Posted 4 Jan 2012 by jujube Post: The URL on the home page and the URL in BOINC manager's Add a Project Wizard are the same URL and that URL should work. If it doesn't work for you then give us some details about what goes wrong or what you see when you try to use that URL. As for there being no work for the past few works, read the thread titled "no more work?".
40) Questions and Answers : Unix/Linux : LHC@home sources available ? (Message 23783) Posted 20 Dec 2011 by jujube Post: This project runs fine on my 2 Kubuntu boxes. This project verifies results by comparing 2 results from the same WU to see if they match. They require an extremely high precision in the results and they have had great difficulty getting results from different platforms to match and verify. After about 5 years of work on the problem they now have results from Linux and Windows matching well and they hope to be able to port the code to OS X in the not too distant future. There is no anticipated delivery date for an OS X app, there isn't even an anticipated date by which they can begin to anticipate a date for delivery. Be aware that the project's sole developer is retired. He doesn't earn a dime from his work at this project and his primary goal at this time is to document and publish how he is able to get results from Linux and Windows to agree with the degree of precision he has achieved. It is a remarkable achievement and it deserves to be documented and published. After that job is finished he hopes to begin porting to OS X. I think I've summarized his concerns and intentions fairly accurately but if you want to check it out for yourself then find Eric Macintosh's posts and read them. I don't recall any official word regarding the source being made available but IMHO they probably won't release it because they want strict control over the task results to ensure they match.

Previous 20 · Next 20

LHC@home