Posts by Igor Zacharov

61) Message boards : Number crunching : Daily quota (Message 23103) Posted 18 Sep 2011 by Igor Zacharov Post: On popular request I have increased the quotas on LHC@HOME 1.0 in config.xml: < <daily_result_quota> 20 </daily_result_quota> WAS --- > <daily_result_quota> 40 </daily_result_quota> NEW 74,75c74,75 < <max_wus_to_send> 5 </max_wus_to_send> WAS < <max_wus_in_progress> 3 </max_wus_in_progress> WAS --- > <max_wus_to_send> 10 </max_wus_to_send> NEW > <max_wus_in_progress> 10 </max_wus_in_progress> NEW Let's see what happens. Any other suggestion of what can be improved?
62) Message boards : Number crunching : No start tag! (Message 23063) Posted 16 Sep 2011 by Igor Zacharov Post: >> EDIT: Ignore...I detached/reattached to project and now all is well. We have changed the URLs and we have adjusted various files in the system. Result is that old values, which are cached in the client do not work any more. This is the case in particular for the scheduler, which gives all sorts of errors. If there are problems, please re-attach to project using the correct URL: http://lhcathomeclassic.cern.ch/sixtrack Please, let us know if the above doesn't help.
63) Message boards : Number crunching : Status of LHC@home (Message 23059) Posted 16 Sep 2011 by Igor Zacharov Post: Friends, we are back in a working condition. Thank you for your patience. If you still experience problems getting work/sending results, please re-attach to http://lhcathomeclassic.cern.ch/sixtrack As we have been restoring operations, we have cleaned the data base. In particular we have touched the teams table to eliminate old errors. Please, report to us any inconsistencies that you see. Our priorities for the coming days/weeks is to run LHC science with your help! The results we have from the run last week are very interesting. In addition, we have found some differences between Intel and AMD processing. We will report back to you when we finish the analysis. From admin, we will focus on updating the project web pages. I expect some initial instabilities, however should not affect the work distribution and results. Thank you.
64) Message boards : Number crunching : Status of LHC@home (Message 23008) Posted 13 Sep 2011 by Igor Zacharov Post: I should comment on the Server Status Page issue. There is no secret in this and there was no attempt to restrict it. It is just some kind of permission problem which I was not able to find so far. It should work and will be on top of the list to fix after the move to the new machine tomorrow.
65) Message boards : Number crunching : Daily quota (Message 22992) Posted 10 Sep 2011 by Igor Zacharov Post: We have crunched quite a few jobs now. Thank you very much for your effort! The rejection factor is small, but it may be worthwhile to understand the numerical cases behind it. We may need to take a pause in submitting new studies until this is understood. It may take us 3 or 4 days. Another change - we should move machines for lhcathomeclassic on wednesday or thursday and also fix some small data base inconsistencies. If you have noticed problems, this is time to report it. I know about some team-leader mismatches, some host-user mismatches, the missing statistics on the lhcathomeclassic/sixtrack page. Anything else?
66) Message boards : Number crunching : Correct project URL? (Message 22977) Posted 9 Sep 2011 by Igor Zacharov Post: the correct URL for LHC@HOME 1.0 aka CLASSIC is http://lhcathomeclassic.cern.ch/sixtrack If you see boinc01 in the URL path, it is the machine name and we have a plan to move to a different machine next week or so. The intention is to keep the URL stated above and physical location of the server will move. More over, the plan is to make some "load balancing" between different machines, such that machines may change on a frequent basis, once we have this machinery setup. Of course, the project URL should be stable.
67) Message boards : Number crunching : Status of LHC@home (Message 22968) Posted 9 Sep 2011 by Igor Zacharov Post: It took some debugging time, but now the validation should work much better. I have increased the number of results/day, so it should send some work now.
68) Message boards : Number crunching : A lot of mismatching results! (Message 22947) Posted 7 Sep 2011 by Igor Zacharov Post: The validation of sixtrack fails for reason not related to numerical comparison. We are still debugging this. Sometimes the validator fails to open the sixtrack output files, while the files are in the upload directory. My guess is that it runs out of memory, or it may be related to the virtual memory setup in which we run boinc. We have prepared a new fresh machine to transfer the sixtrack operations. In the mean time, I will trim the daily quota to 2/day and max wus to send to 1. If our server simply cannot chaw all the results that come in, this will give it some relief and # valid results will increase.
69) Message boards : Number crunching : ? (Message 22940) Posted 7 Sep 2011 by Igor Zacharov Post: I tried to fix this problem. Can you report what you see now? Thanks.
70) Message boards : Number crunching : ? (Message 22921) Posted 7 Sep 2011 by Igor Zacharov Post: We have started a new major study with sixtrack (LHC@HOME 1.0) and I hope there is enough electricity out there! I have left the old study to drain the queues a little bit, so some people have experience the lack of work for a few hours. But now it is in full swing again. The study is based also on the new executable version 530.8 for Linux and Windows. If you happen to check and it is not this version, pls write to me. (the error message about the scheduler.log can be ignored - this is on my side).
71) Message boards : Number crunching : Status of LHC@home (Message 22909) Posted 6 Sep 2011 by Igor Zacharov Post: you are right, of course - we could be much more open about our actions. However, for us, this move was a risky undertaking and we did not know ourselves exactly what it will take to move the service to CERN. Specifically for the URL: the story is not over, there might be another change to be able to facilitate a reliable multi-machine operation. However, this time, as we have the communication with the community going - we will give all info ahead of time and invite comments. By this opportunity - there will be a change in sixtrack executable version tonight. Hopefully, it will be transparent to all of you, as the system will "discover" the new version and will distribute it together with the new work. We expect from the new version (Windows and Linux) a better numerical stability. We have tuned it in the last few days with your formidable help.
72) Message boards : Number crunching : Cannot update Community Preferences (Message 22884) Posted 5 Sep 2011 by Igor Zacharov Post: I have fixed it now. Thank you for your patience!
73) Message boards : Number crunching : Cannot update Community Preferences (Message 22871) Posted 5 Sep 2011 by Igor Zacharov Post: Avatars are now present, but I still cannot get the forums to work as they should. There is some illusive data base error, I suppose which also doesn't allow the community preferences to be edited. Sorry for this. Still tryting to hunt it down.
74) Message boards : LHC@home Science : Linux application finaly here? (Message 22869) Posted 5 Sep 2011 by Igor Zacharov Post: we have both, windows and linux application for the sixtrack program. 97% of all resources are on windows and this work is flowing. Your windows machines should also have work going. I did mention in one of my posts that linux is available, but did not stress this fact. At this time we stil compare the results between windows and linux machines. One of our tasks with sixtrack at this stage is to understand the difference in results we may get from different cpus/architectures/OS, going for bit level equality between them.
75) Message boards : Cafe LHC : Validation Errors (Message 22860) Posted 3 Sep 2011 by Igor Zacharov Post: around mid day today there seem to be some errors on the result assimilation. I have traced some failing validations to that period. To validate, we compare numerical results from several runs. I have taken few examples of the files where the validation status=failed and made the numerical comparison. The results are absolutely identical. Therefore, I'm sure if we run the validator on these results again they will pass. Thank you for reporting this. At this stage we look carefully at the numerics. The assimilation errors started at 13.30 MET and stopped at about 17.00 MET. After that, validation seems to proceed successfully.
76) Message boards : LHC@home Science : Two announcements (Message 22859) Posted 3 Sep 2011 by Igor Zacharov Post: Sänger, I do realize that some glitches happened when transfering the team tables. They are now stable, i.e. we do not plan any admin changes now. If there are still errors in them, please correct by hand. Sorry for the confusion. Dealing with a long running project is more difficult then creating a new one. In what we do we are motivated by science. We have to get results and most of the work so far was dedicated to tuning SixTrack.
77) Message boards : Number crunching : Status of LHC@home (Message 22856) Posted 3 Sep 2011 by Igor Zacharov Post: Dear Friends, the LHC@HOME project has been split in 2 tracks that present themselves as different projects in BOINC: LHC@HOME 1.0 is the classic SixTrack application which is used to study the Beam dynamics at the CERN Collider. We continue with the study and improve the program, which is now available in Linux and Windows. We appreciate VERY MUCH your dedication to this project and count on your support in the future. It is planned that this program is used for serious work leading to the CERN Collider improvements and will ramp up the wu this month. Per decision of CERN Management, we have moved the Boinc Server to CERN site and hopefully have preserved your credits to carry forward with LHC@HOME 1.0. At this time the project's URL is http://lhcathomeclassic.cern.ch/sixtrack . LHC@HOME 2.0 is a novel approach where the program is encapsulated in Vitrual Machine. Theoretical Physics calculation is prepared in the project called Test4Theory. You are wellcome to join that project when it is open for subscription. Separate web pages will be available for that with the URL: http://lhcathomevm/ I believe, we expect still some changes wrt the URLs and will be evolving the web pages to make them more informative. Thank you very much for your support and patience. Igor Zacharov (LHC@HOME 1.0 Transition team).
78) Message boards : Cafe LHC : test (Message 22855) Posted 3 Sep 2011 by Igor Zacharov Post: test
79) Message boards : Cafe LHC : Two announcements (Message 22853) Posted 3 Sep 2011 by Igor Zacharov Post: test message

Previous 20

LHC@home