Posts by old_user14735

1) Questions and Answers : Windows : What has happened here? (Message 23595) Posted 12 Jul 2006 by old_user14735 Post: Ok, great, thanks folks, you have set my mind at rest. I didn\'t want to think that all that cruching had fallen at the final hurdle.
2) Questions and Answers : Windows : What has happened here? (Message 23592) Posted 12 Jul 2006 by old_user14735 Post: I noticed that my model had finihed crunching a little while ago at which point I had these messages in my log... 2/07/2006 19:21:19\|climateprediction.net\|Sending scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi 12/07/2006 19:21:19\|climateprediction.net\|Reason: To send trickle-up message 12/07/2006 19:21:19\|climateprediction.net\|(not requesting new work or reporting completed tasks) 12/07/2006 19:21:23\|climateprediction.net\|Scheduler request succeeded 12/07/2006 19:22:22\|climateprediction.net\|Started upload of file sulphur_ddcr_000623835_0_5.zip 12/07/2006 19:23:20\|climateprediction.net\|Finished upload of file sulphur_ddcr_000623835_0_5.zip 12/07/2006 19:23:20\|climateprediction.net\|Throughput 44308 bytes/sec 12/07/2006 19:28:09\|\|Rescheduling CPU: application exited 12/07/2006 19:28:09\|climateprediction.net\|Computation for task sulphur_ddcr_000623835_0 finished So far, so good... BOINC showed procesing at 100% and (perhaps foolishly) I decided to click the Update button on the project, expecting it to upload the final data. But I got these messages... 12/07/2006 20:45:08\|climateprediction.net\|Sending scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi 12/07/2006 20:45:08\|climateprediction.net\|Reason: Requested by user 12/07/2006 20:45:08\|climateprediction.net\|Reporting 1 tasks 12/07/2006 20:45:12\|climateprediction.net\|Scheduler request succeeded 12/07/2006 20:45:12\|climateprediction.net\|Message from server: Completed result sulphur_ddcr_000623835_0 refused: this result was never sent (the last one in red). Now looking at my account the work unit shows information from all five phases of the model but the server still thinks the unit is not complete. Unfortunately my BOINC client no longer has a CPDN work unit shown. Has my result vanished into limbo? :-( if so is there anything I can do to convince BOINC to try an upload again?
3) Questions and Answers : Windows : Message: request_reschedule_cpus: project op (Message 13549) Posted 18 Jun 2005 by old_user14735 Post: I believe that recent versions of the BOINC client have an improved job scheduling algorithm which takes account of whether a work unit is likely to fail to meet it's deadline (can't remeber where I read this but on one of the BOINC project sites). I think BOINC will give a late unit preferential processing if it thinks it can save the unit from the deadline, over and above the "normal" processing split you've requested (it will certainly stop downloadnig units if it thinks it has too much work). So what may be happening here is that you've got an Einstein unit nearly at the deadline and CPDN is being held back until BOINC completes that time critical unit.
4) Message boards : Number crunching : Announcement: Database residual problem - misallocated WUs (Message 13547) Posted 18 Jun 2005 by old_user14735 Post: I'm currently processing a misallocated work unit (the work unit name in BOINC is 2vvo_300155986_1) which I'm over 50% of the way through. I don't want to kill it if the science will be useful and I don't particularly care whether I get credits or not for this one. Reading the guidelines it seems like I'd still be better to kill it off if it's being processed by someone else though. However, I'm just a little bit confused about how to tell whether my WU is being processed elsewhere or not... I've looked at at the trickles that are recorded against my account and from that list, clicked on the result ID (864094). This shows the name I'd expect (2vvo_300155986_1). When I click the link for the work unit (568245) this shows me a screen where the work unit name is 2vvo_300155986 : exactly the same as mine except without the _1 suffix at the end of the name. Two hosts are marked as crunching it, neither of them me, but both have crashed with a computing error. So on the face of it I'm better to continue assuming 2vvo_300155986 and 2vvo_300155986_1 are really the same thing. If I click on the host ID from the page for result 864094 I see a machine where the latest trickles are for an entirely different work unit. In fact it's one of the machines that crashed on 2vvo_300155986. So it looks like I ought to carry on with this unit. Is this analysis correct and I should continue with 2vvo_300155986_1 or should I kill if off?
5) Questions and Answers : Windows : Credit allocation has stoppped (Message 13545) Posted 18 Jun 2005 by old_user14735 Post: OK thanks, I have another question about this now, but it more properly belongs on the misallocated WU thread I think, so I'll post there...
6) Questions and Answers : Windows : Credit allocation has stoppped (Message 13486) Posted 16 Jun 2005 by old_user14735 Post: Is there an issue with the CPDN servers and the allocation of BOINC credit at the moment? I\'m on BOINC 4.45 and I\'ve just noticed that I haven\'t received any credit for 20 odd days despite steady crunching during this period including an average of about one and a half trickles a day and a successful phase one upload. Admittedly before that period I had a couple of units die before completion, so in terms of \"benefit to the project\" credits I probably deserved to lose some I\'d been given. The crunching on those work units ultimately went nowhere. But as far as I know CPDN isn\'t harsh (or sophisticated) enough to revoke already granted credit or do anything like wait to see if the unit comes back this time before granting anymore credit. So I just wondered if it was a global problem and as no one else seems to have posted on it recently, maybe it\'s just me....
7) Questions and Answers : Windows : Can I manually flag a work unit as an error? (Message 12802) Posted 23 May 2005 by old_user14735 Post: Got a bit of a problem with my current work stream. I had a catastrophic failure of CPDN / BOINC overnight recently. I'm not blaiming BOINC or CPDN for this at all - I know what caused the problem and it was an application I am developing which I left running overnight and which I've now realised was grabbing loads more resources than it should have until it killed the machine (and took BOINC and CPDN with it). I won't be doing that again as I've figured out why my program did this. In fact this was a bad scenario becaue after the failure in the middle of the night BOINC tried to start several new CPDN units with each one crashing immediately until my daily quota was exceeded (good thing that test is present in BOINC). But I have a residual problem with BOINC & CPDN and it's just that the two original CPDN units which were on my machine at the time of the failure (one downloaded in advance by BOINC) and one that was 75%finished are still shown on my account as "in progress" which presumably means they won't be resent until next year. I guess it'd be friendlier to the project to flag them up as failed now as teh deadline is so far in advance. Is there any way I can do this? (I'm on BOINC 4.43 but the units no longer show in my work list)
8) Questions and Answers : Windows : BOINC 4.40 fails to suspend a project when switching (Message 12679) Posted 19 May 2005 by old_user14735 Post: 'fraid I don't know about the scheduler problems (haven't experienced them), but I've been keeping up with the latest BIONC dev versions and CPDN trickles which weren't working for me for a while now seem to be fine since I upgraded to 4.42 yesterday.
9) Questions and Answers : Windows : Warning : BOINC 4.27 screen saver crashes CPDN (Message 11801) Posted 13 Apr 2005 by old_user14735 Post: > Hi David > I just posted some comments <a> href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=2379"> > here,</a> where another person has a problem with the same Intel controller. > > No real answer, just some notes for you to consider. > > Les > Thanks Les. I've added some more information to that thread. I've now upgraded to BOINC 4.30 and I'll try the driver upgrade mentioned there, but with the state of BOINC at the moment, I'm still not going to risk displaying graphics until I've got my latest CPDN work unit safely uploaded.
10) Questions and Answers : Windows : error when I click on (Message 11799) Posted 13 Apr 2005 by old_user14735 Post: I've had the same problem with the same graphics controller and BOINC 4.27. I think BOINC is having lots of graphics problems at the moment! You can read more about my investigations on <a href="http://einstein.phys.uwm.edu/forum_thread.php?id=1163">this thread</a> at Einstein@Home. In summary:- I run all the "production" BOINC projects. I upgraded from 4.19 in the hope that I would get screen saver graphics. In 4.19 I got only a generic BOINC graphic which told me that because of the windows password protection on the screen saver it couldn't show project graphics but that this would be fixed in future upgrades. 4.24 didn't fix my graphics which were now blank. Even worse, once on 4.24 my Einstein science work units started crashing consistently and after quite a bit of experimenting it because apparent that graphics in one form or another, either requested directly or when the (not working properly anyway) screen saver was operating caused the problem. I stopped running them in all forms and all was well. I had the same problems with version 4.25 so stopped running graphics again after a quick experiment. One upgrade later at 4.27 I was tempted into another trial to see if the problem was cured. Graphics were better but not present on all projects. Much worse was that this trial caused a catastrophic crash in CPDN. I don't mind losing the occasional Einstein work unit 'cos they don't take so long, but I don't want to lose all the processing time in a CPDN unit. So I've abstained from graphics althogether and I won't try them again until my latest CPDN model completes. That way I won't lose much if I get another failure... However, in the mean time, I have upgraded to the latest dev version of BIONC (4.30) which promises it fixes some problems and I'll install the Intel driver upgrade mentioned by pwillener. I'm still not risking graphics on my machine until my current CPDN model completes! But if you try either the new BOINC or the graphics driver upgrade and your system seems stable (or not!) please post a reply. I'd be interested to know!
11) Questions and Answers : Windows : Warning : BOINC 4.27 screen saver crashes CPDN (Message 11479) Posted 29 Mar 2005 by old_user14735 Post: I've recently upgraded to BOINC 4.27 (the development version) in the hope that this would resolve a long standing problem I've had with the BOINC screen saver trashing Einstein@home results. There's a long thread about that issue on their forum. The new version of BOINC displayed CPDN graphics on my machine successfully in screen saver mode (the first time that's been done) but won't display graphics on reuqest. Shortly after resuming from the screen saver I lost not only my latest Einstein unit but worse my CPDN calculation crashed out too (CPDN had been shown on the screen saver and was running at the time). My machine is a standard HP 9500 with nothing remarkable in the way of graphics (an "out of the box" Intel 82865G graphics controller) running XP professional with service pack 1. It's probably not advisable to run the BOINC graphics if you upgrade to 4.27.
12) Questions and Answers : Windows : Decommissioning a machine (Message 6263) Posted 20 Nov 2004 by old_user14735 Post: Thanks for all your help folks. I'm trying the "Leave At Least" technique. I believe I've now successfully convinced the machine in question to leave 500 GB, taken it back off line and restored my preferences to 'proper' values for my other BOINC machine. Now I just have to wait for the model finish... I wonder why BOINC holds these preferences centrally? Seems to me that the resource limits a user needs to set, more properly belong at the level of an individual machine and would be better set by the client and read by the server, rather than the other way around. If preferences could be set on a machine by machine basis (or maybe overridden on a machine by machine basis if you wanted a centralised "default" for users of many machines) you'd have a more flexible system, better able to cope with issues like this one. Just some idle design thoughts....
13) Questions and Answers : Windows : Decommissioning a machine (Message 6206) Posted 18 Nov 2004 by old_user14735 Post: > The best way to do it would be to reduce <b>Use no more than</b> in your > general preferences so that the disk space available to BOINC is only around > 500MB more than is currently used for your BOINC projects. > Thanks for the suggestion. That sounds like it might work, but won't the upload of the previous model free up as much space (or maybe even more space) than that required for the download of a new model? So unless I reduced the space to less than I'm currently using, I'd have thought it might still be able to take the unit. I don't know what would happen if I throttle disk usage in my general preferences to a level <b>below</b> my current usage... (and I guess I'd also have to consider the impact on other BOINC machines, making sure I set the preference back again before any of them polled the server and picked up the changed preference). One advantage I have with the machine in question is that it is mostly off line so it isn't going to start any uploading until I connect it to the network. This gives me the opportunity to control the exact time of the upload which should help.
14) Questions and Answers : Windows : Decommissioning a machine (Message 6183) Posted 17 Nov 2004 by old_user14735 Post: I\'ve got a work unit approaching phase 3 completion on a machine which I will shortly be unable to use further for BOINC. I\'d like to ensure that the model finishes and uploads the results, but I don\'t want BOINC to go ahead and download a new model on that machine. What\'s the best way to achieve this?