Message boards : Number crunching : the Problem with Boinc
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Profile San Ra-Lu Quonling

Send message
Joined: 24 Oct 04
Posts: 11
Credit: 27,110
RAC: 0
Message 9517 - Posted: 21 Aug 2005, 5:44:10 UTC
Last modified: 21 Aug 2005, 5:53:39 UTC

Boinc has one Major Problem
it counts the Time the Computer is on
and it counts the Time Boinc is running...

Well normaly was shutting down Boinc if LHC had no work!

Untill the Day i regotniced that i got lesser and lesser WUs to crunsh, cuz the System told my my Computer is not Working 100% for LHC thats a big lol...

sorry but i must say that if i have no work from LHC why should i let Boinc running, and waste my recurces, LHC is running on my Computer everytime if work is there, the Problem is the dam Boinc System..!

now i know, let Boinc run without work, is only good to push the program in that way to think its working 100% but it isnt..

cuz there is no work, well in some kind of thinking for Natural Recurces and Energy saving, i dont understand why i MUST let the Program run cuz the Major uploadside is without work, only to bring Boinc to that state to give me more WU´s for those empty WU Offline Times, every time u shut down boinc cuz there is no work reduces the WU Count u get in Future...

That happens more to users with older Accounts then to users with a fresh and new account, i remember as i start LHC i got 16 WU´s at once cuz my Computer is be capable of to work them in Time, but the last time i only had 3-4 cuz i shut down Boinc ....

thats somekind of sad...

i know i could use some other Programs for crunshing
but i dont like them, the Global weather ist just to big for my Diskspace
Einstein@home is just silly cuz they want theyr WU´s in no time, and if u go on Weekend Vaccation u gain no Points cuz the WU´s ar to old lol....
and much more then LHC it sends annoying messages that the shared work is to low for making wus in time, so i got no work from LHC and form Einstein lol...

thats why i droped all other Programs and Run only LHC
its the Program i like most, but its somekind of silly
if u have a PC and u want share ur worktime and the silly Program told u
that ur not be capable to make WU´s in Time, how do the Program know how long i let my PC on... just cuz the dam count of Vaccation and Networkoffline times ?

This is one more thing why i dont like statistiks...
they are only good for mass problems, but they are mostly wrong for single Probelms....

having shared Power for Work and nobody wants cuz a dam Statistik Funktion just sux...
ID: 9517 · Report as offensive     Reply Quote
Profile verstapp

Send message
Joined: 28 Sep 04
Posts: 22
Credit: 59,075
RAC: 0
Message 9519 - Posted: 21 Aug 2005, 10:38:58 UTC

You can attach to another project so as to do meaningful work while there is no work from LHC. One of the reasons BOINC was invented was to make it easy to switch between projects when one is down for whatever reason.
ID: 9519 · Report as offensive     Reply Quote
Profile Contact
Avatar

Send message
Joined: 17 Sep 04
Posts: 45
Credit: 444,800
RAC: 15
Message 9521 - Posted: 21 Aug 2005, 14:33:23 UTC - in response to Message 9517.  

<blockquote>
sorry but i must say that if i have no work from LHC why should i let Boinc running, and waste my recurces, LHC is running on my Computer everytime if work is there, the Problem is the dam Boinc System..!</blockquote>

I think your complaint is valid. You have pointed out a flaw.
But to only be attached to a project with intermittent work is a rather unique situation and maybe not easily foreseen. Actually now that I see the problem I can’t even imagine the fix.

Try running BOINC 4.19 when LHC is live again. It might work better for your situation.

ID: 9521 · Report as offensive     Reply Quote
Colin Porter

Send message
Joined: 14 Jul 05
Posts: 35
Credit: 71,636
RAC: 0
Message 9522 - Posted: 21 Aug 2005, 15:26:08 UTC - in response to Message 9517.  

<blockquote>
Einstein@home is just silly cuz they want theyr WU´s in no time,. </blockquote>

Einstein has changed their deadline now to 14 days, so may be worth reconsidering.

Personally I prefer LHC but when it's out of work I crunch for Protein

http://predictor.scripps.edu/

Hope you find something to suit.



It's not the speed, but the quality - Until I get a faster computer.
ID: 9522 · Report as offensive     Reply Quote
Profile Nightbird

Send message
Joined: 13 Jul 05
Posts: 55
Credit: 41,230
RAC: 0
Message 9525 - Posted: 21 Aug 2005, 16:58:24 UTC - in response to Message 9522.  

<blockquote><blockquote>
Einstein@home is just silly cuz they want theyr WU´s in no time,. </blockquote>

Einstein has changed their deadline now to 14 days, so may be worth reconsidering.

Personally I prefer LHC but when it's out of work I crunch for Protein

http://predictor.scripps.edu/

Hope you find something to suit.

</blockquote>
Well since the S4 (it seems) with Einstein the deadline is now to 7 days.


Do you want to get banned for 31 years, your account and credits deleted at a Boinc project ? Predictor@home is your best choice.
ID: 9525 · Report as offensive     Reply Quote
Desti

Send message
Joined: 16 Jul 05
Posts: 84
Credit: 1,875,851
RAC: 0
Message 9526 - Posted: 21 Aug 2005, 17:12:24 UTC - in response to Message 9525.  
Last modified: 21 Aug 2005, 17:12:39 UTC

<blockquote>
Well since the S4 (it seems) with Einstein the deadline is now to 7 days.
</blockquote>

My current Einstein@home WU is fresh until 3rd September. :-)
Linux Users Everywhere @ BOINC
[url=http://lhcathome.cern.ch/team_display.php?teamid=717]
ID: 9526 · Report as offensive     Reply Quote
Betting Slip

Send message
Joined: 17 Sep 04
Posts: 41
Credit: 27,497
RAC: 0
Message 9528 - Posted: 21 Aug 2005, 17:36:56 UTC - in response to Message 9526.  


My current Einstein@home WU is fresh until 3rd September. :-)</blockquote>



Well I've downloaded new ones and they are 7 Days expire 28/8 or for yanks 8/28
ID: 9528 · Report as offensive     Reply Quote
Profile Nightbird

Send message
Joined: 13 Jul 05
Posts: 55
Credit: 41,230
RAC: 0
Message 9529 - Posted: 21 Aug 2005, 17:57:18 UTC - in response to Message 9528.  

<blockquote>

Well I've downloaded new ones and they are 7 Days expire 28/8 or for yanks 8/28</blockquote>
Like me

Do you want to get banned for 31 years, your account and credits deleted at a Boinc project ? Predictor@home is your best choice.
ID: 9529 · Report as offensive     Reply Quote
Profile Nightbird

Send message
Joined: 13 Jul 05
Posts: 55
Credit: 41,230
RAC: 0
Message 9531 - Posted: 21 Aug 2005, 19:20:15 UTC

Ok, 14 days for Einstein
"We're starting to issue new workunits (your first server got some of them, in fact) which have a 2-week deadline" (Bruce Allen)

Do you want to get banned for 31 years, your account and credits deleted at a Boinc project ? Predictor@home is your best choice.
ID: 9531 · Report as offensive     Reply Quote
Profile Paul D. Buck

Send message
Joined: 2 Sep 04
Posts: 545
Credit: 148,912
RAC: 0
Message 9549 - Posted: 22 Aug 2005, 13:46:45 UTC - in response to Message 9528.  

<blockquote>
Well I've downloaded new ones and they are 7 Days expire 28/8 or for yanks 8/28</blockquote>
For sailors it is 2005-08-28 or 20050828
ID: 9549 · Report as offensive     Reply Quote
Warped

Send message
Joined: 18 Sep 04
Posts: 40
Credit: 60,176
RAC: 0
Message 9558 - Posted: 22 Aug 2005, 18:33:51 UTC

Try CPDN at:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/index.php

There should be enough to keep you going for the best part of a year on only one work unit.
Warped

ID: 9558 · Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 2 Sep 04
Posts: 165
Credit: 146,925
RAC: 0
Message 9564 - Posted: 22 Aug 2005, 21:02:06 UTC - in response to Message 9525.  

<blockquote><blockquote><blockquote>
Einstein@home is just silly cuz they want theyr WU´s in no time,. </blockquote>

Einstein has changed their deadline now to 14 days, so may be worth reconsidering.

Personally I prefer LHC but when it's out of work I crunch for Protein

http://predictor.scripps.edu/

Hope you find something to suit.

</blockquote>
Well since the S4 (it seems) with Einstein the deadline is now to 7 days.
</blockquote>
The old 7 day deadline WUs from Einstein have not had time to completely flush through the system yet as the change was made less than a week ago. Once a WU is generated, all copies of it get the same duration. All of the 7 day WUs should be gone within a couple more weeks.


BOINC WIKI
ID: 9564 · Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 2 Sep 04
Posts: 165
Credit: 146,925
RAC: 0
Message 9565 - Posted: 22 Aug 2005, 21:05:52 UTC

As to the original problem of BOINC counting the time that it is up. The design decision was to not force computers to be on all of the time. A computer that runs BOINC only 8 hours a day 5 days a week should not be getting the same amount of work as one that is on 24/7.

The problem self corrects if you leave the computer on for a while running BOINC. The key is to attach toseveral projects so that you are not out of work.

If you prefer one project over another use the resource shares to set this. 4.45 and later will cope with extreme variations in resource shares, such ass 100000 to 0.0000001 for example.


BOINC WIKI
ID: 9565 · Report as offensive     Reply Quote
Betting Slip

Send message
Joined: 17 Sep 04
Posts: 41
Credit: 27,497
RAC: 0
Message 9566 - Posted: 22 Aug 2005, 21:08:09 UTC - in response to Message 9564.  

I can't see any reason why for the VAST majority the deadline should be longer than 4 days!
ID: 9566 · Report as offensive     Reply Quote
Profile Ageless
Avatar

Send message
Joined: 18 Sep 04
Posts: 143
Credit: 27,645
RAC: 0
Message 9575 - Posted: 22 Aug 2005, 23:36:58 UTC - in response to Message 9566.  

<blockquote>I can't see any reason why for the VAST majority the deadline should be longer than 4 days!</blockquote>
Not everyone runs Pentium 4s and not everyone runs their computers 24/7. Also not everyone has an always on ADSL/Cable connection.

Thus longer deadlines. To accomodate those that want a bit more time.
Jord

BOINC FAQ Service
ID: 9575 · Report as offensive     Reply Quote
Betting Slip

Send message
Joined: 17 Sep 04
Posts: 41
Credit: 27,497
RAC: 0
Message 9579 - Posted: 23 Aug 2005, 9:17:09 UTC - in response to Message 9575.  

<blockquote><blockquote>I can't see any reason why for the VAST majority the deadline should be longer than 4 days!</blockquote>
Not everyone runs Pentium 4s and not everyone runs their computers 24/7. Also not everyone has an always on ADSL/Cable connection.

Thus longer deadlines. To accomodate those that want a bit more time. </blockquote>



4 Days is a bit more time.
ID: 9579 · Report as offensive     Reply Quote
Desti

Send message
Joined: 16 Jul 05
Posts: 84
Credit: 1,875,851
RAC: 0
Message 9580 - Posted: 23 Aug 2005, 9:25:57 UTC - in response to Message 9579.  
Last modified: 23 Aug 2005, 9:33:41 UTC

IIRC orbit@home will have some workunits with deadline below 24h, but these will be crunched in minutes, not hours.

But for most project 4 days will be too short, just a simple example:

An average PC needs 8 hours for one Einstein@home WU, if it is running only 3 hours every two days, you need 5 days until the WU is finished and returned. But most project are downloading two WU's at the same time, so you need atleast 10 days to finish your jobs.
ID: 9580 · Report as offensive     Reply Quote
Profile Chrulle

Send message
Joined: 27 Jul 04
Posts: 182
Credit: 1,880
RAC: 0
Message 9581 - Posted: 23 Aug 2005, 11:23:11 UTC


Deadline is maybe the wrong name for the thing. Actually it should be reissuetime or something like that. When the deadline is reached and not enough answers for a qourum has returned a new result is created and sent to a new machine. The reissue is counted from the time that the job is submitted, which is why we need a longer deadline. The job may be stuck for some time on the server before it is even downloaded.


Chrulle
Research Assistant &amp; Ex-LHC@home developer
Niels Bohr Institute
ID: 9581 · Report as offensive     Reply Quote
Profile Contact
Avatar

Send message
Joined: 17 Sep 04
Posts: 45
Credit: 444,800
RAC: 15
Message 9585 - Posted: 23 Aug 2005, 14:21:47 UTC - in response to Message 9565.  

<blockquote>The design decision was to not force computers to be on all of the time. A computer that runs BOINC only 8 hours a day 5 days a week should not be getting the same amount of work as one that is on 24/7.</blockquote>
Yes. But San Ra-Lu Quonling's original complaint is that he is crunching 24/7 but only wants to participate in LHC. Why should his 24/7 status be diminished just because LHC is not sending work.
ID: 9585 · Report as offensive     Reply Quote
Profile adrianxw

Send message
Joined: 29 Sep 04
Posts: 179
Credit: 673,019
RAC: 0
Message 9586 - Posted: 23 Aug 2005, 15:17:36 UTC
Last modified: 23 Aug 2005, 15:18:48 UTC

I can kind of see the point and agree with that. If someone, for their own reasons, only wishes to crunch a project which has periods of inactivity, why should the cruncher not have the option of saving power in such situations. After all, we are supposed to be saving energy are we not.

This had not occurred to me before. I guess because I run multiple projects on machines that are running 24/7 anyway, it never happened.

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

ID: 9586 · Report as offensive     Reply Quote
1 · 2 · Next

Message boards : Number crunching : the Problem with Boinc


©2021 CERN