Message boards : Number crunching : Extreme low credits
Author | Message |
---|---|
Quibus Send message Joined: 20 Sep 11 Posts: 4 Credit: 101,051 RAC: 0 |
For me it's never been about the credits, but lately I'm getting about 0.25 for a 25,000 sec. job, anybody knows what this is about? These were the most recent tasks: ABD_20111117_H3_1uz0_ProteinInterfaceDesign_17Nov2011_35452_323 ABC_20111111_H3_1s8n_ProteinInterfaceDesign_11Nov2011_35335_274 Other taks don't have this problem, like Aug20_needle_11start_h2tail_latA_left_SAVE_ALL_OUT__35349_98420 where I got about 44 credit for 8,000 secs of work. |
Quibus Send message Joined: 20 Sep 11 Posts: 4 Credit: 101,051 RAC: 0 |
rm_crs4_20111123_cross_LigDock_fix_2WOR_SAVE_ALL_OUT_35573_2223 Success - Done - 8,540.96 - 50.79 - 0.00 And now 0 credits in spite of success, I don't understand what the problem is... |
dcdc Send message Joined: 3 Nov 05 Posts: 1832 Credit: 119,821,902 RAC: 15,180 |
The tasks are all being validated with 0 credit at the moment. Don't worry about it - I'm sure the team will run a script to assign the correct credit to all of the validated tasks when they can. In the mean-time, carry on crunching ;) |
dondrusco Send message Joined: 2 Jan 07 Posts: 3 Credit: 4,772,623 RAC: 0 |
I tried to set WU run time to 12h (43200s). I have completed successfully one task only. Others finished in ~90 minutes with zero credit... Going back to default (3 hours) and reset project... TaskID WU Id 466738522 425814945 Over Success Done 5,836.20 40.97 0.00 466736890 423842161 Over Success Done 43,123.16 302.74 299.28 466732700 425804666 Over Success Done 6,733.73 47.27 0.00 |
dcdc Send message Joined: 3 Nov 05 Posts: 1832 Credit: 119,821,902 RAC: 15,180 |
I tried to set WU run time to 12h (43200s). I have completed successfully one task only. Others finished in ~90 minutes with zero credit... Going back to default (3 hours) and reset project... There is no need to reset the project and it will make absolutely no difference - the lack of credit being granted is at the server end, not the user end. As stated before, I'm sure valid tasks will get assigned credit at some point and there is no reason to change anything at the user's end in the mean-time. |
Chilean Send message Joined: 16 Oct 05 Posts: 711 Credit: 26,694,507 RAC: 0 |
|
Chilean Send message Joined: 16 Oct 05 Posts: 711 Credit: 26,694,507 RAC: 0 |
|
Quibus Send message Joined: 20 Sep 11 Posts: 4 Credit: 101,051 RAC: 0 |
The tasks are all being validated with 0 credit at the moment. Don't worry about it - I'm sure the team will run a script to assign the correct credit to all of the validated tasks when they can. In the mean-time, carry on crunching ;) Thanks for the info; you're right, it has been credited now. The two with 0.25 credit haven't changed though. Rosetta is the only project I encounter these problems, with several computing errors and such. Is this a Rosetta thing or should I take a look at my settings? |
dcdc Send message Joined: 3 Nov 05 Posts: 1832 Credit: 119,821,902 RAC: 15,180 |
The tasks are all being validated with 0 credit at the moment. Don't worry about it - I'm sure the team will run a script to assign the correct credit to all of the validated tasks when they can. In the mean-time, carry on crunching ;) It's a Rosetta thing. They have lots of people submitting lots of tasks to the server and occasionally a batch has issues that aren't caught pre-release. They're getting tidied up now though so no need for any user action. If you're pushing to advance things like the bakerlab team are, then you're gonna break a few things along the way. It's a small price to pay ;) |
Cutchet Salvador Send message Joined: 1 Feb 10 Posts: 17 Credit: 10,690,439 RAC: 0 |
Unfortunately, the correct WU validated to 0 now have gone on to error. Regrettably nobody of RH opens the mouth to give some type of explanation, as always.... Sad! |
Menace Send message Joined: 4 Nov 05 Posts: 2 Credit: 167,345 RAC: 0 |
I too am getting annoyed by this. I have not received any zero credits, but have a large backlog of un-validated units. Where are the explanations, please? We're donating CPU time (which is money, however you look at it), so it's reasonable to be told what's going on. I've noticed similar 'can't be bothered to tell anyone what's happening' currently at SETI@Home too. I won't get new work for either, and will withdraw from BOINC when what I have is complete. |
Quibus Send message Joined: 20 Sep 11 Posts: 4 Credit: 101,051 RAC: 0 |
I'll be a patient little boy then ;). I don't mind if the server goes offline or things don't go perfect (Seti and Milkyway sure do have their issues too!), it's mostly not knowing what the problem is or if there's a problem to begin with. Thanks for making this clear. |
Cutchet Salvador Send message Joined: 1 Feb 10 Posts: 17 Credit: 10,690,439 RAC: 0 |
It is not necessary to turn into a small patient child, they already treat us like to such without having to do any effort. That different have problems, it is not any reason so that they show any deference with regard to his collaborators. There is alone a question of good practices and education or simply public relations with regard to whom they collaborate disinterestedly with RH. Greetings |
trigggl Send message Joined: 20 Apr 09 Posts: 4 Credit: 102,177 RAC: 0 |
I'd just be happy to get the credits my computer is claiming. I wonder how calculate how much to actually give. |
dcdc Send message Joined: 3 Nov 05 Posts: 1832 Credit: 119,821,902 RAC: 15,180 |
I'd just be happy to get the credits my computer is claiming. I wonder how calculate how much to actually give. It's explained here: https://boinc.bakerlab.org/rosetta/forum_thread.php?id=2194#24612. Basically, it's the average credit claimed for all previously reported tasks of that type, and credit is assigned per model/decoy*. So if the average claimed credit for tasks of type X is 12 credits, and you submit a task with 10 completed models/decoys in it, then you'll get 120 credits. I see you're running linux - I believe the issue there is that in linux the BOINC benchmarks are higher than they are under Windows, but the crunching ability is similar, so your claimed credit (based on the BOINC benchmark) is high, while your granted credit is in line with everyone else's. *Per model or decoy, not per task - there are multiple models or decoys in each task - the longer your run-time, the more decoys it will package in that task. HTH Danny |
MaW Send message Joined: 25 Jun 10 Posts: 4 Credit: 27,878 RAC: 0 |
Over Success Done 10,713.57 61.47 2.84 regardless of the reason(score calculating method), getting 3~ points for 3 hour calculation is just.. discouraging and ridiculous. No wonder that projects that crunch just to crunch (without any significal scientific purpose of the results) are much more popular with their scoring.. I am aware that those are just wortless digits, but ridiculous remains ridiculous. |
mikey Send message Joined: 5 Jan 06 Posts: 1895 Credit: 9,208,737 RAC: 3,249 |
Over Success Done 10,713.57 61.47 2.84 This is a long standing issue...does a project give more credits thus encouraging more users to crunch for them, increasing their workload and possible overloading their Servers, or does it give just enough credits to maintain a happy balance of influx and outflow of users that a fairly even number of users is maintained? Thus keeping everything in balance, as far as the internet can be balanced, Rosetta has made its choice, we either crunch for it or we don't. They are not interested in 50 million people coming and crunching for them right now. |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
Actually... if you know of 50 million people that are interested... :) MaW, sorry to see the low credit task. I don't see any other signs of trouble in that work unit. It must be related to other problems in the project's validator that have cropped up recently. As mikey refers to, there is a long and emotional history about credits across BOINC projects, and even within projects. Many opinions about how to define "fair". Rosetta has done the best it can to grant credit based upon actual work results achieved. Tasks like that do not occur very often. The "long standing issue" that mikey refers to is how much credits are worth, not a task like yours where credit was not properly granted. I don't see anything to ask you about or to change on your end. If you see more of them, please post links to the specific tasks. Rosetta Moderator: Mod.Sense |
MaW Send message Joined: 25 Jun 10 Posts: 4 Credit: 27,878 RAC: 0 |
Next tasks scored a bit more "normally". Well, I guess I can deal with it, It's not like I've just started crunching Rosetta tasks, already used to it. Just that I started in Rosetta again after few months break today and the scoring on first task shocked me =) Also increased the computing time, no point to go with default if the computer is on long enough to take on twice longer units. Actually... if you know of 50 million people that are interested... :) I know of a lot of people in Boinc that get turned on just by seeing high score numbers, so maybe not 50 million but it would still draw lots of people here. But if it's like mikey says that, as I understand, "Rosetta gets just as much computing power as it needs at the moment" it's fine. Anyway, thanks for responding, cheers and Happy New Year, I guess =) |
MaW Send message Joined: 25 Jun 10 Posts: 4 Credit: 27,878 RAC: 0 |
Ah, and so it happened again ;/ Nearly 7 credits for 6h task. https://boinc.bakerlab.org/rosetta/result.php?resultid=474074919 Edit: and the next one is fishy as well... https://boinc.bakerlab.org/rosetta/result.php?resultid=474074970 |
Message boards :
Number crunching :
Extreme low credits
©2024 University of Washington
https://www.bakerlab.org