Message boards :
Number crunching :
LTD / STD
Message board moderation
Author | Message |
---|---|
James Nelson Send message Joined: 23 Mar 02 Posts: 381 Credit: 4,806,382 RAC: 0 |
can someone please explain what the short term and long term debt numbers mean. I have one machine that has stoped getting work for two projects. I thought that it was just trying to make up some debt but its been about 12 days now. Ive been monetering the debt values but cant make sence of them. Is it the LTD or the STD that needs to go down and how far. BTW I did look at the WIKI and couldnt find an explaination. |
Mike Bader Send message Joined: 18 May 99 Posts: 231 Credit: 20,366,214 RAC: 33 |
can someone please explain what the short term and long term debt numbers mean. http://boinc-doc.net/boinc-wiki/index.php?title=Short_Term_Debt Mike Bader BOINC V7.16.5 http://setiathome.berkeley.edu/team_join_form.php?id=5 - Join Our International Team [img]http://boinc.mundayweb.com/one/stats.php? |
James Nelson Send message Joined: 23 Mar 02 Posts: 381 Credit: 4,806,382 RAC: 0 |
can someone please explain what the short term and long term debt numbers mean. yes but what does that mean, that discription is worse than sterio instructions. |
Landroval Send message Joined: 7 Oct 01 Posts: 188 Credit: 2,098,881 RAC: 1 |
yes but what does that mean, that discription is worse than sterio instructions. When a project has the CPU, its long-term debt decreases (goes toward the negative direction); when a project is waiting for someone else, its debt increases (goes toward the positive direction). The debt is the amount of CPU time owed *to* that project. Each time the scheduler switches between projects, debt is recalculated for all projects, and the numbers are then adjusted so that the total of all debt across all projects is 0. If a project gets more than its resource share (earliest-deadline-first to make deadline, other projects not having work, etc) its LTD can go below 0. If its LTD goes farther below 0 than 1 time slice (i.e. below -3600 for the default settings), then that project is prevented from downloading new work. Work it's already downloaded will continue to be processed as usual--round-robin if it's not under deadline pressure, earliest-deadline-first if it is. The project can download new work when its long-term debt returns to something close to 0. Example: 3 projects, equal resource share. Each should get the CPU about 1/3 of the time. In normal round-robin mode each project gets the CPU one hour each, in rotation. You could think of it this way: When it gets the CPU, Project A runs for 1 hour. Because it was only 'entitled' to 1/3 of that hour, it owes 2/3 of an hour--1/3 to each of the other 2 projects. The next hour, project B runs, and A sits idle--1/3 of an hour repaid. The next hour C runs, and again A sits idle, repaying 1/3 hour of long-term debt. At the end of 3 hours, A has run for 1 hour, accumulating 2/3 hour of debt, then paying it back over the next 2 hours, and it's A's turn again. Now suppose A had a lot of work on hand that was under deadline pressure and ran for 60 hours straight. A now owes 40 hours (2/3 of 60 hours) long term debt. A will not be allowed to download more work until it's paid that debt off. Every hour that A sits idle pays off 1/3 hour of long-term debt. So A's debt will start out much less than 0, because A has used more than its share, and will gradually return to 0 as the debt is paid off. B and C, meanwhile, will have large positive debts--they're both owed quite a bit, because they haven't had the CPU at all for 60 hours. Their debts will gradually come down as they divide the CPU between them 50/50 for a while. If the projects have different resource shares the numbers get awkward, but the same principle applies. If the 2 projects still have LTD that's much less than 0, they're still paying off debt. Hope this helps, Brian If you think education is expensive, try ignorance. |
River~~ Send message Joined: 25 Dec 05 Posts: 28 Credit: 87 RAC: 0 |
STD decides which project next gets to run. The highest number goes next - but this is updated just before the client makes its decsion so sometimes it doesn't look that way. LTD decides which projects are allowed to download more work. Projects that are more than a little negative will only download work if a cpu is idle and all projects with higher LTD fail to send work. For the exact meaning of 'a little negative' you will have to refer to the wiki. LTD is not the only thing that can stop work being downloaded. If the client adds a cacheful of time to the existing mix and decides that something will run out of deadline, then it will not download work from any project whatever the LTDs say. You can spot this from the message 'work fetch suspended, machine overcommitted' or something like, followed eventually we hope by 'resuming work fetch'. Also a project cannot download work if 'No More Work' has been set, or if that *project* is suspended. However suspending a result does not prevent that project downloading more work - often in direct response to your suspend command! Be careful when suspending to know if you are in the work tab (still allows fetch) or the project tab (doesn't allow fetch). Finally, and puzzling to many newcomers, if the button says 'suspend' the project/result is allowed to run. If the button says 'resume' it is not allowed to run. Reason: the label tells you what the button does next time it is clicked, it does not tell you the current state. Ditto Allow work fetch / No more work - tells you what the button will do, which is the opposite of the current state of the client. Hope at least one of the above helps you solve your problem. Or maybe there is no problem and the client just wants to get a bit more space to the next deadline before getting new work. Good luck River~~ |
Lee Carre Send message Joined: 21 Apr 00 Posts: 1459 Credit: 58,485 RAC: 0 |
when BOINC does a "check" the project with the highest (positive) STD is the one that will be run next the project with the highest (positive) LTD is the one that it will download work from next, if no work is available, it'll try each project in order of decreasing LTD is this the kind of answer you were looking for? |
Tern Send message Joined: 4 Dec 03 Posts: 1122 Credit: 13,376,822 RAC: 44 |
I just read the Wiki page, and _I_ can't come up with any better way to say it. It's perfectly clear to me... If you would give us the LTD numbers from _your_ host, we could tell you specifically what's going on. Maybe a concrete example would help. |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
can someone please explain what the short term and long term debt numbers mean. Let me try: Short term debt controls how BOINC schedules work already downloaded to your computer. The most positive project is most likely to run. Long term debt controls downloads. Positive projects will be downloaded, negative projects won't. The most common exception: if all the "positive" projects are out of work, negative projects will download one work unit. If you crunch several projects, scheduling is much easier if you keep a small cache. If you keep a large cache, you're more likely to go through periods where BOINC will not download work for one project or another. |
James Nelson Send message Joined: 23 Mar 02 Posts: 381 Credit: 4,806,382 RAC: 0 |
when BOINC does a "check" the project with the highest (positive) STD is the one that will be run next I'd like to thank everyone who replied. I guess its the LTD that needs to get closer to zero. I'll let it continue to run the way that it is for a while longer, don't really want to reset my debt values, or to reset the other two projects, this has just been going on a long time and I was getting concerned. so as long as my LTD keeps progressing toward zero Ill let it keep doing its thing. |
Landroval Send message Joined: 7 Oct 01 Posts: 188 Credit: 2,098,881 RAC: 1 |
I'd like to thank everyone who replied. I guess its the LTD that needs to get closer to zero. I'll let it continue to run the way that it is for a while longer, don't really want to reset my debt values, or to reset the other two projects, this has just been going on a long time and I was getting concerned. so as long as my LTD keeps progressing toward zero Ill let it keep doing its thing. If you're running Windows, BoincDV is a handy little utility that lets you view the short-term and long-term debt for each project. And you're right, "let it run and let the scheduler do its thing" is probably the best course here. Trying to micromanage the scheduler usually just makes its flaky behavior worse. Happy crunching! |
James Nelson Send message Joined: 23 Mar 02 Posts: 381 Credit: 4,806,382 RAC: 0 |
[/quote] If you're running Windows, BoincDV is a handy little utility that lets you view the short-term and long-term debt for each project. And you're right, "let it run and let the scheduler do its thing" is probably the best course here. Trying to micromanage the scheduler usually just makes its flaky behavior worse. Happy crunching! [/quote] thankyou I all ready ues boincdv. BTW one of the projects just started back up today, hopefully the other will start soon. My numbers are starting to slip and I want to maintain my rank or go up if possible. |
Steve Cressman Send message Joined: 6 Jun 02 Posts: 583 Credit: 65,644 RAC: 0 |
Then can you explain why projects with NO work continue to get more negative. It is only supposed to get more negetive when that project is actually doing work. For example Orbit and Burp are not giving out work but they continue to get more and more negative everyday. But they still keep asking for work everyday with very very negitive numbers. This is all in reference to LTD values. Also uFluids LTD has never been positive yet it crunches uFluid every day with positive STD numbers. This not a complaint. And I won't complain about this because my resource shares are being honored. Just curious as to why I am seeing it function in this maner. Boinc v5.2.13 PRJ: SETI@HOME STD: -1140.926312 LTD: 33371.523739 RSRC: 3 ------------------------------- PRJ: CLIMATEPRE STD: 0.000000 LTD: -14766.208571 RSRC: 0 ------------------------------- PRJ: EINSTEIN@H STD: 933.382780 LTD: 43623.143533 RSRC: 10 ------------------------------- PRJ: BURP STD: 0.000000 LTD: -14247.532231 RSRC: 10 ------------------------------- PRJ: ORBIT@HOME STD: 0.000000 LTD: -14766.208571 RSRC: 5 ------------------------------- PRJ: UFLUIDS STD: 24.291376 LTD: -34378.551180 RSRC: 3 ------------------------------- PRJ: SETI@HOMEB STD: 195.759214 LTD: 1163.833280 RSRC: 2.500000 ------------------------------- 98SE XP2500+ @ 2.1 GHz Boinc v5.8.8 And God said"Let there be light."But then the program crashed because he was trying to access the 'light' property of a NULL universe pointer. |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
Then can you explain why projects with NO work continue to get more negative. It is only supposed to get more negetive when that project is actually doing work. For example Orbit and Burp are not giving out work but they continue to get more and more negative everyday. But they still keep asking for work everyday with very very negitive numbers. This is all in reference to LTD values. LTD only comes into effect in determining which projects to download work from. If there is work from that project on the system, it will be crunched no matter what the LTD is. The STD determines which WU to crunch next, and has no effect on what to download. I noticed that Orbit@Home and Climate Prediction have exactly the same LTD, and that BURP is pretty close. It is my guess that you have reset or detached from a project with negative LTD that caused these projects to go to negative LTD in response. If you reset a project, its LTD is set to 0 momentarily (it is immediately moved away from 0 again as the LTDs are shifted to make the mean be 0). BOINC WIKI |
Steve Cressman Send message Joined: 6 Jun 02 Posts: 583 Credit: 65,644 RAC: 0 |
Yes John , I did reset all values to zero 5 or 6 days ago so I could see what the scheduler is currently doing and to get the projects that are not doing anything back to zero LTD. My understanding is that a project needs to run (crunch work)in order for the debt to go negative. Climate has resource share of zero so is never doing work but still gets more and more negative. That goes for Orbit as well since it has no winblows app yet and Burp is on temporary hiatus. Also my other point was that Ufluids LTD is always negative and Ufluid does d/l work while the LTD is negative. Which I don't think it should do unless that is the only place to get work from. Hope this clarifies the parts I am trying to understand. Also I knew that STD calls the work to be done and that LTD decides where to get it from. Thanks John for your response. But I'm Still Curious :) EDIT: uFluid just dowloaded more work just now and the debt values are PRJ: UFLUIDS STD: -913.416446 LTD: -28545.081769 RSRC: 3 98SE XP2500+ @ 2.1 GHz Boinc v5.8.8 And God said"Let there be light."But then the program crashed because he was trying to access the 'light' property of a NULL universe pointer. |
River~~ Send message Joined: 25 Dec 05 Posts: 28 Credit: 87 RAC: 0 |
... The rule is slightly more complex than that. Small negative values do not stop d/l, the LTD has to be more negative than a certain amount. (From memory the cut-off value is something like minus the cache size in seconds, but look in the wiki to confirm please) |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
... Negative of the project switch time (default 60 minutes). BOINC WIKI |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
Yes John , I did reset all values to zero 5 or 6 days ago so I could see what the scheduler is currently doing and to get the projects that are not doing anything back to zero LTD. My understanding is that a project needs to run (crunch work)in order for the debt to go negative. Climate has resource share of zero so is never doing work but still gets more and more negative. That goes for Orbit as well since it has no winblows app yet and Burp is on temporary hiatus. What version of BOINC? BOINC WIKI |
itenginerd Send message Joined: 1 Aug 00 Posts: 37 Credit: 39,905 RAC: 0 |
What version of BOINC? Boinc v5.2.13 He slipped that in an earlier post. Gotta help out... I'm kinda interested to see where this one goes. 8) [edit] I'm blind too--it's in his sig. Gaarrrh. Watch the obvious! 8P [/edit] (j) James |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
YIKES, a resource share of 0? Could you do a suspend instead? The code does divide by resource share all over the place. I suppose that since the code has not faulted, they must all be special cased, but the special cases could be causing trouble. BOINC WIKI |
Steve Cressman Send message Joined: 6 Jun 02 Posts: 583 Credit: 65,644 RAC: 0 |
YIKES, a resource share of 0? Could you do a suspend instead? The code does divide by resource share all over the place. I suppose that since the code has not faulted, they must all be special cased, but the special cases could be causing trouble. Not only is climate's share zero but 'no new work' and 'suspend' are set and verified under project status. REsource share is at zero so that from the manager I won't accidentally request from a project that for now I don't want to run. All the rest of the projects are set to run and get work if they can. So i now know that LTD does not have to be positive to request work because it is tied into the 'Switch between applications every' value. I use an 8 hour switch time right now because of uFluid. So now just the question of why do these ones not proccessing work keep getting more negative? Some time ago John I thought you had fixed the schedular so that this would no longer happen. Burp, Orbit and Climate are the ones I am talking about. It can't be the request for work making the LTD go more negative because Climate never asks for more. So if any thing they should get more and more positive because the other projects that are running would owe a debt to the ones not running. Thanks for your time :) 98SE XP2500+ @ 2.1 GHz Boinc v5.8.8 And God said"Let there be light."But then the program crashed because he was trying to access the 'light' property of a NULL universe pointer. |
©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.