 |
|
|
21)
Message boards :
Number crunching :
Can\'t trickle in
(Message 22491)
Posted 30 Apr 2006 by Nuadormrac Post: This is the message I\'m getting now: 4/29/2006 6:50:03 PM|climateprediction.net|Sending scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi However, and several things. - I did connect to climateprediction.net when I first connected - This is not a new, never before been received WU, as can be seen here. It successfully trickled in several times http://climateapps2.oucs.ox.ac.uk/cpdnboinc/graph_cm3.php?resultid=5106979 - I sure as hell don\'t want to detach or do anything that would force a reset, or cause me to lose this unit I was successfully crunching and trickling in on thus far, without incident... HELP!!!! |
|
22)
Message boards :
Number crunching :
Restoring WU, forcing BOINC CC to forget the WU errored out
(Message 22148)
Posted 17 Apr 2006 by Nuadormrac Post: OK, thx |
|
23)
Message boards :
Number crunching :
Restoring WU, forcing BOINC CC to forget the WU errored out
(Message 22121)
Posted 17 Apr 2006 by Nuadormrac Post: Well, I guess in a way it\'s better it got killed; though thinking something happened locally I really did try to get it to go, and sorta screwed up BOINC with all projects re-setting. Well, I shoulda waited for a reply, and not thought \"oh my gosh, lost all this crunch time, what do I do\", while trying to \"fix\" things? :D Course I also didn\'t fair well on sulpher units in the past, so when a coupled was going good, and then...but couldn\'t understand on a load, if I was going nuts, or what :rofl BTW, on the other board, I can\'t seem to login... When the BOINC servers went down following some network error, I couldn\'t get to this group. So I created an account on the other board, however the confirmation email was never sent to me. Probably the same network problem at the time of creation that prevented our machines from trickling in, or don\'t know. Anyhow, a login shows account is non-existent, incorrect password, or inactive. Select send password, and it shows my email address as in use, but the account as inactive. Not sure how I would go about resolving that... |
|
24)
Message boards :
Number crunching :
Restoring WU, forcing BOINC CC to forget the WU errored out
(Message 22114)
Posted 17 Apr 2006 by Nuadormrac Post: :eek: So the model didn\'t really crash due to a problem at my end, but an auto-reset was forced from the network, or? Was strange, start WU, and on load, dead... |
|
25)
Message boards :
Number crunching :
Restoring WU, forcing BOINC CC to forget the WU errored out
(Message 22110)
Posted 17 Apr 2006 by Nuadormrac Post: Weird, after sometime of crunching LHC (while it was there), I resumed CPDN. I know it got further in the time step before without incident when I last suspended, but it just suddenly failed when loading. So I shut down BOINC, deleted the CPDN folder, and copied from backup. Problem is BOINC still says \"ready to report\". What files do I edit, or whatever, to force BOINC to forget that it completed the WU and was ready to force, and make BOINC resume from backup as if nothing ever happened, with the CC being none the wiser? thx in advance... |
|
26)
Message boards :
Number crunching :
Anyone have experience with CC 5.4.x and climate predictor here?
(Message 22030)
Posted 14 Apr 2006 by Nuadormrac Post: OK, thx... Then I\'m gathering all should be well helping out on the RALPH project, and also allow the CPDN WU to progress. |
|
27)
Message boards :
Number crunching :
Anyone have experience with CC 5.4.x and climate predictor here?
(Message 22025)
Posted 13 Apr 2006 by Nuadormrac Post: OK, here\'s the skinny. I\'m also crunching RALPH (Rosseta Alpha), and they\'re trying to work out the reason the newer Rosseta app keeps crashing on HBLR_* type WUs, a few of the newer types, and also have been trying to work out the \"1% bug\". All seemed fine last weekend (before the new WU types were released), and 4.97 was put up on Rosseta. The new types came out about then and most of the Windows systems started having crashing probs. Didn\'t effect Macs, and seems to be fine on Linux (even on the same machines, and I did check at least here), but where all the bug reports are comming in from). On the main project, user complaints were piling up, and the app was rolled back, with 4.98 Rosseta realling being the same as 4.83... So, in the midst of trying to isolate this, Rom Walton had asked that everyone crunching RALPH to put CC 5.4.1 on their comps, indicating that the older clients do not collect the sort of debugger info they\'re needing to diagnose the cause of these crashes. He also noted that yes it is an alpha project, so a pre-release CC isn\'t beyond reasoning when trying to track down probs in their new science app... Anyhow, I have a couple model about 5% in (just short of it), that had been running happily. I did upgrade the CC to give the Rosseta team the info they need to help hunt down the bug in their new app. However wondering if anyone has tried this on Climate Predictor yet, and if so what they\'re experience between the new beta CC and this project has been. The backup is less then a days (actual crunch time, not calander time) out of date, and running with networking disabled for now, but wondering... |
|
28)
Message boards :
Number crunching :
No Credit - Why?
(Message 22024)
Posted 13 Apr 2006 by Nuadormrac Post: That\'s not a problem with this project per se... That\'s more an issue that your machine isn\'t up to spec. SETI and Einstein have less stringent resource requirements, that\'s why you don\'t see that. When a project puts recommendations on their site for a given model, much like a software company puts it on the box, there is a certain degree of reasoning behind those recommendations. One other thing of note... Having things set to \"leave in memory\" can help wrt the not reaching a checkpoint issue. It is also a normal recommendation for a project such as this. However, and here\'s the unfortunate part, \"leave in memory\", has this stuff paged out, and if you\'re having to page a hell of a lot already, hmm... You\'re committed memory between physical and pagefile will definitely go up keeping all projects in memory... |
|
29)
Message boards :
Number crunching :
Is it worth it...
(Message 22023)
Posted 13 Apr 2006 by Nuadormrac Post: One other thing to note... Looking at the history from slab, to sulpher, to coupled... I almost would gaurentee that the next phase will not be shorter then coupled. If you notice the trend, the WUs are getting larger. So if they go to another phase in a year I\'d expect the current trend to continue with WU sizes... That comp might be done with CPDN if it won\'t let you d/l... Now, this leaves 2 matters. For now, faster computers can pick up much of the slack and still let people complete it in deadline with time to spare for other projects. My Athlon 64 in fact has no seeming deadline trouble project sharing between CPDN with a coupled, seasonal attribution, and some other projects... In time however, as AMD and Intel go multi-core as a means to increase computing power; in part because they can\'t using current manufacturing technology just continue to ramp up clock rates as of old. Conroe, will based on some preliminary benchies give a mighty boost to performance, but like the Athlon 64 it\'s going to be a 64-bit x86 CPU, and will also move in a new direction from Intel, become much more efficient as they also drop the net-burst arch of the P4 altogether. Conroe will also be multi-core like the X2, based on the preview they demonstrated comparing it to an OCed FX60!... Part of the problem is heat dissipation on these newer procs, and part of it is the laws of physics facing companies such as Intel and AMD... One of the main methods of dealing with heat dissipation was a die shrink. However the silicon atom has size, and eventually one just can\'t shrink any further without hitting the sub-atomic, or going to a smaller atom such as carbon (which shares many similar properties to Si on the periodic table). The other thing is that die shrinks had posed a bit problematic, difficult to get in working order. There\'s most definitely a reason that Intel hadn\'t continued ramping clock rates to a doubling every 18 months or so for awhile. Long term, places such as Sandia National Labs (which is local to Albuq. NM here) are looking for replacements to the silicon based semi-conductor, which won\'t share the same technological limits; however such research takes time. In the mean time, multi-core and other such introductions does seem to be the way these companies are going on to compete, as they\'re comming closer to certain limits. Of course 2 cores can do more, however it\'s the same deal as with SMP... |
|
30)
Message boards :
Cafe CPDN :
Just an FYI: Seasonal attribution webserver having probs
(Message 21465)
Posted 20 Mar 2006 by Nuadormrac Post: Yeah, not sure when it\'ll come back up. Yesterday, the loads were slow, and then nothing... BOINC stats is now reporting seasonal attribution as offline http://www.boincstats.com/ Status schedulers: with all other projects showing green. Wondering if maybe the server crashed under whatever load was bringing it to such a slowdown or? Networking might also be an issue, though if that project and this is on the same network pipe, this project is up and running without incident... |
|
31)
Message boards :
Cafe CPDN :
Just an FYI: Seasonal attribution webserver having probs
(Message 21440)
Posted 20 Mar 2006 by Nuadormrac Post: At first I wasn\'t sure if it was just at home or what, but connections were slow, and pages off the site not always loading... However I\'m now at uni (University of New Mexico here), and over their several T3 lines which are hardly being used given it\'s a Sunday night and hardly anyone is here) I\'m actually having even greater trouble with half loaded pages and timeouts... Thought I\'d alert someone, to, well something going on with either the servers or Internet connectivity to them... The connection to the CPDN board here has conversly been fine all day without issue from either place... |
|
32)
Message boards :
Number crunching :
A question of credit.
(Message 21439)
Posted 20 Mar 2006 by Nuadormrac Post: Thanks a bunch... That was it, the overlay... Anyhow I\'m getting 1.92 s/TS so I guess that\'s good... I\'m guessing doing the math that tricles are now 25k a part?, or every January per model year? I guess I\'ll find out... |
|
33)
Message boards :
Number crunching :
A question of credit.
(Message 21436)
Posted 20 Mar 2006 by Nuadormrac Post: By TCM, do you mean the new 5.06 coupled unit? I got one of these units and had a few questions... Perhaps a mod wouldn\'t mind putting a sticky on a thread containing this info, so it\'s out there for people who look. Up to them though... My questions about the new 5.06 coupled units is this: - How often do these units have save points? - What is the credit per trickle (is this the 266/trickle mentioned)? - Hitting 8 on my keyboard doesn\'t show the countdown to save point, nor are the stats in coupled (though in seasonal they do display) shown for s/TS or TS completed. Huh? - If one disables networking while away, will it handle attempts to trickle in with networking disabled very well? The first one I ask, because it\'s always good to know when it\'s had a savepoint, so one knows the best time to shutdown/backup while losing the minimum amount of crunch time... The last I ask, because last night I came home from dinner to an unpleasent surprise. Kaspersky 6.0 beta (release candidate actually) totally locked up Windows due to a beta bug, and my entire BOINC cache got waisted. Seasonal already reported in while I was away, so nothing I could do; it automatically reported a failure. If one disabled networking while away for a length of time, to give one\'s self a chance to shut down/restart from the last save point or backup (if possible), will the model have problems with no networking when it attempts to get out? thx in advance... |
|
34)
Message boards :
Number crunching :
Sulpher unit shows up in my profile that I never downloaded
(Message 21435)
Posted 20 Mar 2006 by Nuadormrac Post: OK, thx... I was taken aback by this discrepency of a unit showing up in my acct, and didn\'t like the idea they\'d have to wait a year to find out. Guess they do have error handling for it then. BTW, seeing this I did reconnect, and got one of the new combined WUs... I\'ll save pertinent questions on it for another thread. |
|
35)
Message boards :
Number crunching :
Sulpher unit shows up in my profile that I never downloaded
(Message 21433)
Posted 20 Mar 2006 by Nuadormrac Post: Not sure what to make of this... Was just checking the account and it shows a sulpher unit sent to me in January... Only problem is my PC never got it, deadline set to next year... Huh? I abviously can\'t allert the servers to this discrepency, so what should I do if anything? |
|
36)
Message boards :
Number crunching :
Downloaded sulphur 4.22, not coupled 5.08
(Message 21431)
Posted 19 Mar 2006 by Nuadormrac Post: I sorta wish they had the option in their BOINC project specifics preference page to select which type of units one wanted to run (ala what world community grid allows in their preferences page), so in cases where one could run one more successfully then another, they chould tell it chose that... Not sure why sulpher has proved a bit problematic, where the more intensive CPDN seasonal has proven a better match for my comp... I thought it was supposed to be the other way around, but it\'s the case. Course sulpher is also experimental by the suggestion, so there could still be a software bug or the more dreaded system specific compatibility issue in there, that only makes itself a nuisance in some situations... Seasonal WUs still help these people out though, so I had been able to return to some CPDNish crunching after a lapse :) |
|
37)
Message boards :
Number crunching :
Average s/TS of 8.42! Any point in me continuing?
(Message 21426)
Posted 19 Mar 2006 by Nuadormrac Post: You have one year to complete the unit in, and someone calculated it to taking 128 days... You\'re well within deadline so I wouldn\'t worry about it... True, I saw much lower crunch times (I say saw, because after several sulfer units failed I\'m of the conclusion that sulfer 4.22 hates my A64 and will wait for 4.23)... In my case, and if you could double the RAM to 1 GB you might try the CPDN seasonal attribution project, which is running much more successfully on my A64 and also has an estimated 28 day runtime. But from what some have indicated there\'s a reason it\'s recommended to have 1 GB of RAM, as some who have less have claimed their swapfile gets thrashed... Can\'t comment as I do have a GB so... However, it\'s only when you can\'t complete in a year you should be concerned... Your comp can complete this with 24/7 operation and still have > 6 months time to spare in actual run time... |
|
38)
Message boards :
Number crunching :
Problems with sulphur_4.22_windows_intel.exe?
(Message 19461)
Posted 20 Jan 2006 by Nuadormrac Post: Yes, both of those settings are set to yes. Also, my computer was idle and I wasn\'t present when this happened (might not have been at home). I just got back, checked the WUs, and noticed that climate predictor was no longer at the top of the list. A further inquiry showed it as .51% completed or so (whereas before, it was closer to 10% on the WU that was being worked on). |
|
39)
Message boards :
Number crunching :
Problems with sulphur_4.22_windows_intel.exe?
(Message 19438)
Posted 19 Jan 2006 by Nuadormrac Post: I\'m getting failed WUs...and tbh have been trying to honor the mention that we should complete a WU after signing up. I can\'t get these to complete. OK, on a former install, after I added this project and did some other things I did have a bit of a prob with kernel paged pool reaching the 128 MB limit, and then well Windows crashing... Since then I\'ve upgraded the RAM/replaced the old and that isn\'t happening. It\'s possible, the old mobo went, that maybe it took the RAM with it somewhat... OK, so I re-connected the project and http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=1626073 1/18/2006 3:40:39 AM|climateprediction.net|Sending scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi 1/18/2006 3:40:39 AM|climateprediction.net|Reason: To send trickle-up message 1/18/2006 3:40:39 AM|climateprediction.net|Note: not requesting new work or reporting results 1/18/2006 3:40:44 AM|climateprediction.net|Scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi succeeded 1/18/2006 3:46:04 AM|climateprediction.net|Sending scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi 1/18/2006 3:46:04 AM|climateprediction.net|Reason: To send trickle-up message 1/18/2006 3:46:04 AM|climateprediction.net|Note: not requesting new work or reporting results 1/18/2006 3:46:09 AM|climateprediction.net|Scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi succeeded 1/18/2006 3:47:59 AM||request_reschedule_cpus: process exited 1/18/2006 3:47:59 AM|climateprediction.net|Computation for result sulphur_iukz_000879443_0 finished BTW, this is with an Athlon 64, not an Intel. But it\'s one of the newer Athlon 64 cores using the Venice core, which unlike the previous Athlon 64 cores includes SSE3 instructions (well 11 of them that don\'t relate to hyper-threading). As the newer A64 cores adds this from the Intel instruction set, don\'t know if BOINC can make use of this, or if this could somehow be related or not... |
|
40)
Message boards :
Number crunching :
Project is down?
(Message 16659)
Posted 18 Oct 2005 by Nuadormrac Post: Oh, so much for posting late at night and having not checked the dates. Post did seem somewhat close to the top of the posts list :D Guess I should have started another thread then. Anyhow, the first trickle did get in, but the confirmation email hasn\'t arrived though on another attempt it has both my email address and user name as taken (by me), and the option to email the admins still shows the address as undeliverable. Guess I\'ll need an admin to recover a registered account in case all the BOINC boards and all go down again... |
©2024 climateprediction.net