Grid computing center

log in

Advanced search

Message boards : News : Grid computing center

1 · 2 · 3 · 4 . . . 7 · Next
Author Message
Profile [AF>WildWildWest] Sebastien
     
Dictator
Avatar
Send message
Joined: 28 Mar 10
Posts: 2678
Credit: 513,735
RAC: 95
Total hours: 1,427,438
Message 5366 - Posted: 9 Jul 2017, 19:06:53 UTC

Because of some participants who abused by multiplying clients which run only Grid computing center, workunits from those clients will be invalid. No hours will be counted.

mmonnin
       
Send message
Joined: 22 Aug 16
Posts: 417
Credit: 1,928,033
RAC: 705
Total hours: 9,131,146
Message 5367 - Posted: 9 Jul 2017, 19:43:54 UTC - in response to Message 5366.

Because of some participants who abused by multiplying clients which run only Grid computing center, workunits from those clients will be invalid. No hours will be counted.


GCC admin encourages multiple clients, has publicly said so and they do it themselves.

Profile JayPi
     
Send message
Joined: 8 Oct 11
Posts: 12
Credit: 5,494,396
RAC: 142
Total hours: 10,671,965
Message 5369 - Posted: 10 Jul 2017, 16:10:48 UTC

This will happen also on physical Hosts running only GCC if other projects are suspended or have no WUs!
For example this Host: http://wuprop.boinc-af.org/results.php?hostid=86962

mmonnin
       
Send message
Joined: 22 Aug 16
Posts: 417
Credit: 1,928,033
RAC: 705
Total hours: 9,131,146
Message 5370 - Posted: 10 Jul 2017, 16:57:49 UTC
Last modified: 10 Jul 2017, 16:59:03 UTC

I don't have any invalids but I have some clients with only GCC/WUProp. My GCC/Radioactive hours have varied quite a bit since the announcement though. Down, back up and now down again about 80 hours each direction. Might have to setup each client to run a a single small CPU task every 6 hours if the admin wants to play games.

Profile aendgraend
       
Send message
Joined: 30 Mar 10
Posts: 5
Credit: 1,061,896
RAC: 103
Total hours: 2,518,290
Message 5371 - Posted: 10 Jul 2017, 21:19:52 UTC - in response to Message 5366.
Last modified: 10 Jul 2017, 21:25:29 UTC

Because of some participants who abused by multiplying clients which run only Grid computing center, workunits from those clients will be invalid. No hours will be counted.


I'm sorry to having to say that, but in my Opinion it's an absolutely stupid decision to invalid correctly processed Workunits and not counting the Hours. Nothing else.

You lead your project to absurdity if real processed work isn't counted anymore. That hasn't anything to do with multipying hosts. If one sets up a hosts, be it virtual or not, one invests in Hardware and/or Electrocity Costs to process the Work. In Addition, the Project itself encourages the participants to do so.

So, as I said above, stupid decision, and in hindsight you'll come to pretty much the same insight.

HAL9000
 
Avatar
Send message
Joined: 14 Dec 11
Posts: 33
Credit: 6,479,481
RAC: 319
Total hours: 9,528,885
Message 5374 - Posted: 11 Jul 2017, 11:06:14 UTC - in response to Message 5370.

I don't have any invalids but I have some clients with only GCC/WUProp. My GCC/Radioactive hours have varied quite a bit since the announcement though. Down, back up and now down again about 80 hours each direction. Might have to setup each client to run a a single small CPU task every 6 hours if the admin wants to play games.

I added CPU projects to some of my hosts that previously were GCC only and their WUProp tasks are still being marked as invalid after 18 hours of running CPU work.
I guess they are being marked invalid by a method other than not having a CPU project?

mmonnin
       
Send message
Joined: 22 Aug 16
Posts: 417
Credit: 1,928,033
RAC: 705
Total hours: 9,131,146
Message 5375 - Posted: 11 Jul 2017, 11:45:09 UTC - in response to Message 5374.
Last modified: 11 Jul 2017, 12:12:08 UTC

I don't have any invalids but I have some clients with only GCC/WUProp. My GCC/Radioactive hours have varied quite a bit since the announcement though. Down, back up and now down again about 80 hours each direction. Might have to setup each client to run a a single small CPU task every 6 hours if the admin wants to play games.

I added CPU projects to some of my hosts that previously were GCC only and their WUProp tasks are still being marked as invalid after 18 hours of running CPU work.
I guess they are being marked invalid by a method other than not having a CPU project?


The one I checked says:
Erreur assignation taille application

Google translates to:
Application size error

Maybe reset the project? Del the exe and download another copy? Or copy from another instance.

2.1mil GCC hours. That's crazy.

My hours have been cut in half and invalid but show no error. Poor choice admin. Maybe I'll just run 32 clients with 1 task each and end up getting more points. Looks like other NCI apps don't count.

Also you realize this is killing hours for other apps as well?

Profile Brian Uitti
   
Avatar
Send message
Joined: 27 Jan 11
Posts: 13
Credit: 208,842
RAC: 0
Total hours: 416,688
Message 5376 - Posted: 11 Jul 2017, 17:09:17 UTC - in response to Message 5366.

Because of some participants who abused by multiplying clients which run only Grid computing center, workunits from those clients will be invalid. No hours will be counted.


I must be missing something..

As I see it, the only real problem is if you're running multiple sessions of WUprop on the same system, and counting the same work done twice.. regardless of which projects are running.. no need to pick out Grid Computing Center specifically. Running multiple sessions of any project should be getting more work done, and should be valid

On top of that, wouldn't WUprop only count the work done within that virtual machine? .. so everything done should count? .. right?

Correct me if I'm wrong.

I look at "cheating" as being both unethical and pointless. I would hope that most people here agree. I only hope that this project properly counts actual work being done and reports it accurately in a way that's meaningful and practical.

// Brian
____________
Expanding the edge of Science.

k4m1k4z3
   
Send message
Joined: 3 Jul 11
Posts: 2
Credit: 576,592
RAC: 24
Total hours: 2,177,946
Message 5377 - Posted: 11 Jul 2017, 17:28:24 UTC

I have machines that only run 1 instance of BOINC and are only running GCC and wuprop... and you are telling me that this is not valid?

You know that people run multiple clients on the same PC for CPU intensive applications in order to gain more hours? The invalidating of work where GCC was the only project is pointless.

mmonnin
       
Send message
Joined: 22 Aug 16
Posts: 417
Credit: 1,928,033
RAC: 705
Total hours: 9,131,146
Message 5378 - Posted: 11 Jul 2017, 17:29:49 UTC - in response to Message 5376.

Because of some participants who abused by multiplying clients which run only Grid computing center, workunits from those clients will be invalid. No hours will be counted.


I must be missing something..

As I see it, the only real problem is if you're running multiple sessions of WUprop on the same system, and counting the same work done twice.. regardless of which projects are running.. no need to pick out Grid Computing Center specifically. Running multiple sessions of any project should be getting more work done, and should be valid

On top of that, wouldn't WUprop only count the work done within that virtual machine? .. so everything done should count? .. right?

Correct me if I'm wrong.

I look at "cheating" as being both unethical and pointless. I would hope that most people here agree. I only hope that this project properly counts actual work being done and reports it accurately in a way that's meaningful and practical.

// Brian


WUProo won't count hours twice for the same client. Each client has its down gui_rpc_port and WUProp needs to use that for work/hours to validate in the 1st place. Even with all that setup correctly the admin has decided to give us the middle finger.

GCC/Monkeys is a Project to guessed what words a monkey would press if they were sat down in front of a computer and start mashing buttons, to my knowledge. Each task is separate, returns different results, validates and is allowed by GCC. I could see the admin here complaining about QCN getting credit multiple times at the same site. It's not really gathering anything different. No one is even running those projects so who cares. GCC is like any other CPU project, it's just NCI and can be run multiple times on the same computer.

HAL9000
 
Avatar
Send message
Joined: 14 Dec 11
Posts: 33
Credit: 6,479,481
RAC: 319
Total hours: 9,528,885
Message 5380 - Posted: 12 Jul 2017, 0:19:00 UTC - in response to Message 5375.
Last modified: 12 Jul 2017, 0:20:01 UTC

I don't have any invalids but I have some clients with only GCC/WUProp. My GCC/Radioactive hours have varied quite a bit since the announcement though. Down, back up and now down again about 80 hours each direction. Might have to setup each client to run a a single small CPU task every 6 hours if the admin wants to play games.

I added CPU projects to some of my hosts that previously were GCC only and their WUProp tasks are still being marked as invalid after 18 hours of running CPU work.
I guess they are being marked invalid by a method other than not having a CPU project?


The one I checked says:
Erreur assignation taille application

Google translates to:
Application size error

Maybe reset the project? Del the exe and download another copy? Or copy from another instance.

2.1mil GCC hours. That's crazy.

My hours have been cut in half and invalid but show no error. Poor choice admin. Maybe I'll just run 32 clients with 1 task each and end up getting more points. Looks like other NCI apps don't count.

Also you realize this is killing hours for other apps as well?

I figured I would try detach/attach for WUprop. Then see if those hosts running CPU apps would start having valid tasks. I may need to force the BOINC clients to generate a new CPID before reattaching. Not sure yet.

I've been a fan of the Infinite monkey theorem before the BOINC project for it existed. So far I have put a bit more than 2.1 million hours into the monkey apps. It is closer to about 8.5 million hours to date.

I currently have several dual Xeon E5-2670 systems and was hoping to have a few of them run the monkey CPU apps. Once the BOINC bug preventing NCI and non-NCI apps from the same project is fixed and the project is able to distribute the CPU apps alongside the NCI apps. I wonder if those apps will have their time counted?

mmonnin
       
Send message
Joined: 22 Aug 16
Posts: 417
Credit: 1,928,033
RAC: 705
Total hours: 9,131,146
Message 5381 - Posted: 12 Jul 2017, 1:02:04 UTC - in response to Message 5380.

I don't have any invalids but I have some clients with only GCC/WUProp. My GCC/Radioactive hours have varied quite a bit since the announcement though. Down, back up and now down again about 80 hours each direction. Might have to setup each client to run a a single small CPU task every 6 hours if the admin wants to play games.

I added CPU projects to some of my hosts that previously were GCC only and their WUProp tasks are still being marked as invalid after 18 hours of running CPU work.
I guess they are being marked invalid by a method other than not having a CPU project?


The one I checked says:
Erreur assignation taille application

Google translates to:
Application size error

Maybe reset the project? Del the exe and download another copy? Or copy from another instance.

2.1mil GCC hours. That's crazy.

My hours have been cut in half and invalid but show no error. Poor choice admin. Maybe I'll just run 32 clients with 1 task each and end up getting more points. Looks like other NCI apps don't count.

Also you realize this is killing hours for other apps as well?

I figured I would try detach/attach for WUprop. Then see if those hosts running CPU apps would start having valid tasks. I may need to force the BOINC clients to generate a new CPID before reattaching. Not sure yet.

I've been a fan of the Infinite monkey theorem before the BOINC project for it existed. So far I have put a bit more than 2.1 million hours into the monkey apps. It is closer to about 8.5 million hours to date.

I currently have several dual Xeon E5-2670 systems and was hoping to have a few of them run the monkey CPU apps. Once the BOINC bug preventing NCI and non-NCI apps from the same project is fixed and the project is able to distribute the CPU apps alongside the NCI apps. I wonder if those apps will have their time counted?


I have a E5-2670 v1 2P system as well. I had been running 2 clients at a full 32 threads with some other NCI only clients. I just dropped the 2nd client down about half threads and started running the other clients with 1 thread. Drug Discovery came back which has quick units to run on the other clients. WUProp tasks have begun validating and GCC hours are going back up.

Cruncher Pete
       
Send message
Joined: 28 Mar 10
Posts: 20
Credit: 2,107,041
RAC: 0
Total hours: 25,291,622
Message 5388 - Posted: 12 Jul 2017, 23:14:23 UTC

Please clarify this for I do not believe that you are penalizing me for using all cores of my machines to crunch Grid Computing Center WU's only on my machines. I have paid considerable sums of money to purchase my Xeon CPU's, some running 76 cores, some 56 and others 24. Are you telling me to cut back using these machines utilizing them to their full capacity? I do not beleive this is what is intended to crunch BOINC projects. If my machines crunch these projects legitimately than why am I not getting the hours for it? GCU have not mentioned anything to me or others that we are crunching too many WU's for their liking. Perhaps you might care to suggest where I might be able to purchase single core CPU's.
____________

Cruncher Pete
       
Send message
Joined: 28 Mar 10
Posts: 20
Credit: 2,107,041
RAC: 0
Total hours: 25,291,622
Message 5389 - Posted: 13 Jul 2017, 0:18:08 UTC

The Project Administrator Of GridComputing Center {BOINC@Poland]goofyx stated the following on his website on 12 July:

"It's strage... maybe WUProp change their rules, but it's not fair for 99,9% crunchers.
On GG@H I have maybe 2 user with really masiive rigs for multi boinc but it was for 3 weeks.


ps.: Monkeys has 4 NCI apps so you can get 4*24h=96hour per day per computer."

Whilst you have all the right in the world to grant whatever hours you wish on your project, you are undermining the Gridcomputercenter project by telling their users not to use all their resources on that project alone, thereby favoring other sites.

I ask why do you penalize users with large number of powerfull machines to crunch any BOINC projects.?
____________

HAL9000
 
Avatar
Send message
Joined: 14 Dec 11
Posts: 33
Credit: 6,479,481
RAC: 319
Total hours: 9,528,885
Message 5390 - Posted: 13 Jul 2017, 2:03:35 UTC - in response to Message 5381.

I don't have any invalids but I have some clients with only GCC/WUProp. My GCC/Radioactive hours have varied quite a bit since the announcement though. Down, back up and now down again about 80 hours each direction. Might have to setup each client to run a a single small CPU task every 6 hours if the admin wants to play games.

I added CPU projects to some of my hosts that previously were GCC only and their WUProp tasks are still being marked as invalid after 18 hours of running CPU work.
I guess they are being marked invalid by a method other than not having a CPU project?


The one I checked says:
Erreur assignation taille application

Google translates to:
Application size error

Maybe reset the project? Del the exe and download another copy? Or copy from another instance.

2.1mil GCC hours. That's crazy.

My hours have been cut in half and invalid but show no error. Poor choice admin. Maybe I'll just run 32 clients with 1 task each and end up getting more points. Looks like other NCI apps don't count.

Also you realize this is killing hours for other apps as well?

I figured I would try detach/attach for WUprop. Then see if those hosts running CPU apps would start having valid tasks. I may need to force the BOINC clients to generate a new CPID before reattaching. Not sure yet.

I've been a fan of the Infinite monkey theorem before the BOINC project for it existed. So far I have put a bit more than 2.1 million hours into the monkey apps. It is closer to about 8.5 million hours to date.

I currently have several dual Xeon E5-2670 systems and was hoping to have a few of them run the monkey CPU apps. Once the BOINC bug preventing NCI and non-NCI apps from the same project is fixed and the project is able to distribute the CPU apps alongside the NCI apps. I wonder if those apps will have their time counted?


I have a E5-2670 v1 2P system as well. I had been running 2 clients at a full 32 threads with some other NCI only clients. I just dropped the 2nd client down about half threads and started running the other clients with 1 thread. Drug Discovery came back which has quick units to run on the other clients. WUProp tasks have begun validating and GCC hours are going back up.

At the moment I am limited to running 10 of my dual Xeon E5-2670 systems at once. At least until I have the unused electric dryer circuit moved from the laundry room to the computer loft. I have that scheduled to be done in Sept.
Then I'll probably only keep the systems running flat out when it is cool enough that the 5KW of heat doesn't make my place to warm. I might have to figure out some creative means of getting rid of all that heat.

Cruncher Pete
       
Send message
Joined: 28 Mar 10
Posts: 20
Credit: 2,107,041
RAC: 0
Total hours: 25,291,622
Message 5391 - Posted: 13 Jul 2017, 2:47:28 UTC - in response to Message 5366.
Last modified: 13 Jul 2017, 3:03:33 UTC

Because of some participants who abused by multiplying clients which run only Grid computing center, workunits from those clients will be invalid. No hours will be counted.


I still need clarification that I do not understand. I have a number of Xeon CPU systems and a number of Raspberry Pi machines all running only Grid Computing center. According to your statement they will be invalid and hours will not be counted, because I am only running them on gridcomputer center.. Why am I being disciminated by not giving me credits. A number of other systems run gridcomputing center and other projects and you will credit those? The Admin of GridComputing center states that he is satisfied with 99.9% of users submission, why are you different?
____________

Profile [AF>WildWildWest] Sebastien
     
Dictator
Avatar
Send message
Joined: 28 Mar 10
Posts: 2678
Credit: 513,735
RAC: 95
Total hours: 1,427,438
Message 5392 - Posted: 13 Jul 2017, 3:04:12 UTC - in response to Message 5388.

Please clarify this for I do not believe that you are penalizing me for using all cores of my machines to crunch Grid Computing Center WU's only on my machines. I have paid considerable sums of money to purchase my Xeon CPU's, some running 76 cores, some 56 and others 24. Are you telling me to cut back using these machines utilizing them to their full capacity? I do not beleive this is what is intended to crunch BOINC projects. If my machines crunch these projects legitimately than why am I not getting the hours for it? GCU have not mentioned anything to me or others that we are crunching too many WU's for their liking. Perhaps you might care to suggest where I might be able to purchase single core CPU's.

There is no problem if you run one client per host. Hours will be counted.
There was a problem with some users which ran until 100 clients per hosts.

Same users are also cheating by running Radioactive or QCN projects without sensors. So either they stop cheating before next week, either project will be closed.
I can't spent many hours to detect cheaters.
____________

Cruncher Pete
       
Send message
Joined: 28 Mar 10
Posts: 20
Credit: 2,107,041
RAC: 0
Total hours: 25,291,622
Message 5393 - Posted: 13 Jul 2017, 4:08:20 UTC - in response to Message 5392.

Please clarify this for I do not believe that you are penalizing me for using all cores of my machines to crunch Grid Computing Center WU's only on my machines. I have paid considerable sums of money to purchase my Xeon CPU's, some running 76 cores, some 56 and others 24. Are you telling me to cut back using these machines utilizing them to their full capacity? I do not beleive this is what is intended to crunch BOINC projects. If my machines crunch these projects legitimately than why am I not getting the hours for it? GCU have not mentioned anything to me or others that we are crunching too many WU's for their liking. Perhaps you might care to suggest where I might be able to purchase single core CPU's.

There is no problem if you run one client per host. Hours will be counted.
There was a problem with some users which ran until 100 clients per hosts.

Same users are also cheating by running Radioactive or QCN projects without sensors. So either they stop cheating before next week, either project will be closed.
I can't spent many hours to detect cheaters.


I can't accept that statement to be your legitimate reason for dictating to the users what they can or can not crunch. Surely, if the project is happy with 99.9 percent of users submission (as stated by the Admin of gridcomputingcenter two days ago) why do you wish to tell the users not to crunch their projects with all their resources and just limit your input to just one client, one host, otherwise they will not get hours recognition for them. Surely you are blackbanning the project by placing restrictions on their use. As regards your restriction to run one client per host. BOINC Client/Manager allows multiple BOINC Clients to run concurrently. This is to utilize 100% of the CPU. All projects accept this accept you. I will not run my machines capacity at some ..% when I can set it to run a number of Wu's at 100% thereby helping the project. May I suggest that if this is not satisfactory to you than you should take it up with the BOINC Management Team and perhaps restrict all users input to just a single CPU. Backward thinking if I ever saw one. You will do us a favor not to waste our money on better and better equipment and it will also reduce the cost of electricity and waste of time`.

Finally, if you can't spend many hours detecting cheaters, may I suggest that you stop being a policeman and just attend to your own project. Policing and interpretation of what is considered cheating is up to individual projects to determine. They all have their options and means to detect and police their own. May I ask why do you consider cheating by running Radioactive or QCN project without a sensor? why should it be a concern to WuProp?. Surely, it is a project problem and it should be fixed by them. Why are they giving credit in the first place, if users do not meet the requirement. I do not call that cheating. For example, I often run all available BOINC projects and let BOINC Manager handle it. Why are those projects listed and available if they meet special needs. Surely, if I do not meet their requirement they should not download a WU to me and give me credit for completing a WU legitimately. I object being told what to crunch and how much resources I should allocate to a project.
____________

Cruncher Pete
       
Send message
Joined: 28 Mar 10
Posts: 20
Credit: 2,107,041
RAC: 0
Total hours: 25,291,622
Message 5394 - Posted: 13 Jul 2017, 6:05:15 UTC - in response to Message 5393.

Same users are also cheating by running Radioactive or QCN projects without sensors. So either they stop cheating before next week, either project will be closed.
I can't spent many hours to detect cheaters.


This is getting worse and worse. Personally it does not apply to me for I have not crunched any of those projects for a long time. I do posses a sensor for Radioactive and I have never crunched QCN. I do have to question what do you consider the be cheating. I can prove to you that if I legitimately log into Radioactive and my sensor is not connected I will not get credit for it, yet WuProp will grant hours worked for the time. Do you not think that it is your project that is at fault?

If you "can't spent many hours to detect cheaters" than make sure that the people you consider cheaters are in deed cheaters and it is not your projects buggy system.

Please prove to me/us who are the cheaters and stop using the legitimate users by threatening the closure of the project and stop threatening others who are acting within the BOINC software options by exercising their legal rights to use MUltiple BOINC Clients on their machines. It is time that Project Managers recognize just how much money it costs to build a number of machines for BOINC use and the cost of electricity that we need to run those costly machines. Whilst I admit there will likely to be cheaters in he system, I am offended by suggesting that some projects will close because of that, disappointing all of us..
____________

mmonnin
       
Send message
Joined: 22 Aug 16
Posts: 417
Credit: 1,928,033
RAC: 705
Total hours: 9,131,146
Message 5397 - Posted: 13 Jul 2017, 11:35:34 UTC

QCN isn't evening validating legit work any more. I have two tablets setup that should report accurately but I've still never received a point. Zero points for the project in over 30 days.
http://stats.free-dc.org/stats.php?page=proj&proj=qcn

You are blocking all clients with NCI only though. Even if all that was running is GCC, you're still invalidating legit work.

There aren't even enough NCI apps out there to get a single star so it's not like NCI apps can ruin the WUProp star system. Those that have a purple star or really any star got them through CPU and GPU projects.

1 · 2 · 3 · 4 . . . 7 · Next
Post to thread

Message boards : News : Grid computing center


Home | My Account | Message Boards | Results


Copyright © 2024 Sebastien