Not Requesting New Work Units
Message boards :
Number crunching :
Not Requesting New Work Units
Message board moderation
| Author | Message |
|---|---|
|
Daniel Collins Send message Joined: 8 May 05 Posts: 8 Credit: 965,682 RAC: 62 
|
Greetings. I downloaded the newest version of BOINC in the hopes that my BOINC would finally download more packets. I have not received packets in several days. However I have recently turned in more packets (I download 10 days worth at a time). I can connect to the SETI Server without a problem. Please advise. BOINC during its startup has showed something different than it has before for SETI, maybe this might help. This is from the Messages tab on BOINC listed SETI@home as the project: 1/3/2006 5:54:44 PM|SETI@home|Computer ID: 860836; location: home; project prefs: default 1/3/2006 5:54:44 PM||General prefs: from Einstein@Home (last modified 2005-06-24 03:22:28) 1/3/2006 5:54:44 PM||General prefs: no separate prefs for home; using your defaults 1/3/2006 5:54:45 PM||Remote control not allowed; using loopback address Please advise Daniel S. Collins ID: 225491 ·  |
|
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0
|
Greetings. I downloaded the newest version of BOINC in the hopes that my BOINC would finally download more packets. I have not received packets in several days. However I have recently turned in more packets (I download 10 days worth at a time). I can connect to the SETI Server without a problem. Please advise. BOINC during its startup has showed something different than it has before for SETI, maybe this might help. This doesn't say much other than you've got SETI and Einstein running. Without knowing anything else, it is possible that BOINC is going to do most of the work you have in the cache (to make sure it gets completed by the deadlines) and then grab a bunch of SETI and do nothing but SETI for most of the following ten days. Why? You've got a 10 day connection interval, and work with a 14 day deadline. The second connection (in 20 days) is six days past deadline, so BOINC will go to extra lengths to make sure it finishes as much as possible before the next connection. I'd recommend setting the "connect every 'x' interval" to one day, give BOINC time to stabilize (which may take a while) and then slowly increase the interval while you watch how BOINC behaves. ID: 225496 · |
|
Daniel Collins Send message Joined: 8 May 05 Posts: 8 Credit: 965,682 RAC: 62
|
Hmm.... I already have the interval set like that... however I will give it a few days and see if it works. I currently am actively participating in 4 projects... 3 of which are acting like SETI.. the 4th.. Climate Predictor is working fine.. Einstein, SETI, and Predictor all are giving me Not Requesting New Work Units.... But maybe the new BOINC will work but just take time.. I will letcha know... Thanks for the help Daniel S. Collins ID: 225521 · |
|
Daniel Collins Send message Joined: 8 May 05 Posts: 8 Credit: 965,682 RAC: 62
|
Btw.. one more thing... I dont have any work units to work on.. atleast no work units on any besides Climate Predictor which is only set to a 22% share of resources.. just wanted to make sure I told ya that.. chat with ya in a couple of days.. and thanks again Daniel S. Collins ID: 225523 · |
 Landroval Send message Joined: 7 Oct 01 Posts: 188 Credit: 2,098,881 RAC: 1
|
I currently am actively participating in 4 projects... 3 of which are acting like SETI.. the 4th.. Climate Predictor is working fine.. If for some reason they've been getting more than their resource share lately, they'll hold off on downloading more until they've paid off the debt. Could that be it? If so, once one of them has waited long enough, it'll request more work and off you go. Each will come back 'online' one at a time as the long-term debt falls. Sometimes working out the debt takes a while.... Cheers, Brian If you think education is expensive, try ignorance.  
ID: 225526 · |
|
Landroval Send message Joined: 7 Oct 01 Posts: 188 Credit: 2,098,881 RAC: 1
|
Btw.. one more thing... I dont have any work units to work on.. atleast no work units on any besides Climate Predictor which is only set to a 22% share of resources.. So the other projects have 78% of the resources between them? If they've been getting more, for whatever reason, it's possible they all owe some time to Climate. I've got something similar on my machine...it downloads a bunch of Einstein, works round-robin for a while, goes into EDF to finish it off (5-day work queue), then Einstein has to wait a while while CPDN has sole use of the CPU for a day or so until long-term debt evens out. Give it a couple days to straighten out. ID: 225527 · |
Tern  Send message Joined: 4 Dec 03 Posts: 1122 Credit: 13,376,822 RAC: 44
 |
Greetings. I downloaded the newest version of BOINC in the hopes that my BOINC would finally download more packets. I have not received packets in several days. However I have recently turned in more packets (I download 10 days worth at a time). Confused; you later say you have the cache set to 1 day. But then you say you're doing 4 projects, including CPDN, so the _maximum_ allowable cache size to avoid being in EDF mode constantly is 3.5 days, and the recommended is 1. What version of BOINC did you have before 5.2.x? If it was pre-4.45, where there _was_ no scheduler, it wouldn't have cared what your settings were, it would have just rotated projects every hour (per resource share), even if this meant missing deadlines. V4.45 would have not missed deadlines, but could wind up blowing your resource share settings. The newer ones are better at "following your rules", but in doing so, enter EDF mode and refuse to get new work quite often, IF your cache is too large. ID: 225562 · |
John McLeod VII  Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0
|
Btw.. one more thing... I dont have any work units to work on.. atleast no work units on any besides Climate Predictor which is only set to a 22% share of resources.. How much time is estimated till completion of the CPDN result? (use BOINC Manager, NOT BOINC View for this measurement). How long is it to the deadline? What are the time stats (Up time, On time, ...) (available in client_state.xml) What is the Duration Correction Factor of CPDN? (probably 1.0, but just checking). (available in client_state.xml) What is the resource share of CPDN? What is the resource share of the next project in line to ask for work (Highest LTD with no backoff)? (info available in client_state.xml) Estimated wall time = estimated CPU time * DCF * Up time * On time * CPU efficiency * runnable resource share. If the CPDN result is going to take more than 90% of the time till its deadline, the host will be in EDF and NWF because downloading any more work at this time is likely to make the CPDN result late. If the CPDN result is going to take more than 90% of the time till its deadline with its CPU time reduced by (CPDN resource share / (CPDN resource share + Other resource share)) the host will be in NWF mode because downloading any more work could cause the CPDN result to be late.   BOINC WIKI ID: 225569 · |
|
Daniel Collins Send message Joined: 8 May 05 Posts: 8 Credit: 965,682 RAC: 62
|
ok.. let me explain this from the beginning.. to tell you the truth I dont understand some of the technical words being used like CPDN and such.. I have been using BOINC since May of 2005 I think... with no problems after I got it up and started... I have had 4 projects... Seti has 45% of the cpu time set to it... the other 3 compromise 55%... Einstein, SETI, and Predictor no longer requests work units at all.. they are all completed and results has been returned to the respective projects. The only project that I still have data processing is Climate Predictor. Before, BOINC would download lots of packets from SETI, Fewer from Einstein, alot from Predictor, and only 1 from climate predictor.. my computer would be able to finish them all in the time allotted (except for a period of time when my internet went down and went past the expiration time a while back)... I really havnt had much of a problem working with BOINC at all. as of about 2 days ago BOINC will not request work units from SETI (atleast that what is says in BOINC)... I came from seti classic back in May and I do think that BOINC is definately a step in the right direction (for multitasking multiple projects).. I dont understand BOINC much besides its setup and running. I dont understand how to make this work... can you please help me understand what you are talking about when you say CPDN and all the other initialized wording... I currently have my computer set to a 10 day interval.... however, on a daily basis (normally at night) I go ahead and upload all my completed workunits to the respective projects and it downloads more packets. I do this just in case the servers go down and it has come in handy when servers have gone down. Excuse me for not understanding what you were asking ealier. I will be eagerly awaiting a reply... Daniel S. Collins ID: 225627 · |
|
Daniel Collins Send message Joined: 8 May 05 Posts: 8 Credit: 965,682 RAC: 62
|
Btw.. one more thing... I dont have any work units to work on.. atleast no work units on any besides Climate Predictor which is only set to a 22% share of resources.. This sounds like that this might be the answer to my problem.. thank you... I just want to make sure BOINC is working perfectly... here is the resource distribution for my projects... SETI@HOME 44.44% PREDICTOR 11.11% EINSTEIN 22.22% CLIMATEPRE 22.22% ----------------- TOAL 99.99% Climate predictor has 386 hours till completion (like it always has)... Currently it has 36 Hours processed on this packet... 7.64% Normally it never finishes before it stops and sends some result back to Climate Predictor's server. I think this answers questions to the other post too... I read back through it all and think I got a better grasp of what you all were asking. I dont mean to seem like a pain... I just want it working.. <G> but you have put my mind at ease... thank you Daniel S. Collins ID: 225631 · |
|
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0
|
ok.. let me explain this from the beginning.. to tell you the truth I dont understand some of the technical words being used like CPDN and such.. This all makes sense. JM7 is the expert, if you give him the numbers he seeks, he'll explain why it does what it's doing, and when it will ask for more SETI. The rest of us are basically saying, in general, that Predictor is owed work, and the others are being held off until Predictor catches up. By the way, with many projects, you can set "connect every 'x' days" to a fairly low value (like 0.5 days). If one project is down, BOINC will crunch the other three, and when the down project comes back, BOINC will crunch it until it catches up. ID: 225634 · |
|
Tern Send message Joined: 4 Dec 03 Posts: 1122 Credit: 13,376,822 RAC: 44
|
CPDN is ClimatePredictionDotNet... silly abbreviation, but that's what most call it. The basic problem you have is the 10-day cache. JM7 will give you the exact figures for what is going on, if you give him all the data, but in general, when you have a cache that is larger than the "maximum recommended", you will only be working on one project at a time. This is because you are forcing the computer to be in EDF (Earliest Deadline First) mode, instead of "round robin" (alternating every hour), by having too much work on hand. Your current CPDN result is a "slab" model; when you get the next one, it will likely be a "sulphur" model, which will take more like 1000 hours. At that point the problem will become even worse. The scheduler is smart enough to keep you from missing deadlines, and it's also very good at matching your requested resource share, but it has a much easier time doing this if you don't fill the cache with so many results. If you will lower your cache to 3 days, and wait a couple of weeks for your current load to be worked out, and for the new setting to settle down, you'll see it getting work and doing "round robin" scheduling more like you expect. (All projects instead of one at a time.) ID: 225647 · |
|
Paul D. Buck Send message Joined: 19 Jul 00 Posts: 3898 Credit: 1,158,042 RAC: 0
|
ok.. let me explain this from the beginning.. to tell you the truth I dont understand some of the technical words being used like CPDN and such.. Daniel, The Wiki is your friend ... :) Click on any of the links, like the one by my signature to get to the Wiki, but in ANY of the terms or phrases in the search box (use the google one prefereably) and more definitions than you probably want to read. IF, AFTER YOU READ that stuff, it is still not clear. Create a new thread and ask a question. One of the people that "patrols" the boards to offer help will likely give an explanation. But, one of the things that will get you for a bit will be the need to look up all the new words and phrases used until you have a more complete BOINC "vocabulary". ID: 225735 · |
Astro  Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
here is the resource distribution for my projects... Daniel, I went to boincstats.com and typed in your user name, then hit search. I then clicked on the "graph" icon under boinc and from the piechart labelled "credit divided over projects" I see you have: 53.9% Seti 20.7% Einstein 12.8% PPAH (Protein Predictor @ home) 10.2% CPDN (Climate Prediction) 2.5% Sztaki So It seems you're getting the resource share you asked for, now it's just understanding why/how it's getting there. tony PS the scheduler divides up work via wall clock run time NOT total credits, so this pie chart won't exactly match your preferences, and If I had to guess I'd guess you signed up for seti all by it self for a while and that's why the seti credit total exceeds 44%. Also, from the numbers I've shown you that CPDN needs more run time to get back up to 22%  ID: 225736 · |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.