Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
![]() |
World Community Grid Forums
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 1130
|
![]() |
Author |
|
RMau
Cruncher Joined: Feb 6, 2008 Post Count: 44 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Farpung,
----------------------------------------I have tried all of the usual tricks to catch some specific work. And I agree, most likely there just isn't work from DDDT-2 to be had right now. When I decided to try leaving the machined idle, I didn't know that the time interval between attempts to fetch work increased as the idle period increased. So there is a great deal of diminishing return here. The idle period between communication attempts will very quickly grow to equal the time it would take to crunch a WU from another project. I won't leave the machines idle for too long. It is just an experiment since all other attempts at grabbing a DDDT-2 WU, or two, have not been successful. ![]() |
||
|
mclaver
Veteran Cruncher Joined: Dec 19, 2005 Post Count: 566 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
I've heard the pleas for patience. But if this is to be thought of as an intermittent project, and not a project released to be crunched, then the admins should put it in either Beta or Intermittent status. The science is, in the end, what we're all hoping to contribute to. But WCG constructed the 'Ladder of Badges' in part to keep folks involved in crunching by giving them something tangible for their efforts. So, in a community of 'all for the science', it is disconcerting to see what appears to be a distribution model that favors certain high performance machines, rather than distributing the 'science' to all participants. I like the 'beta or intermittent' characterization for what is happening with DDDT-2. I can get behind that knowing that the beta process does have requirements for things like WU return times. The sooner the beta is back, the sooner people can look at the results and make needed corrections, etc. So, pull DDDT-2 back and call it what it is, a beta. When the bugs are worked out and it is ready for EVERYONE to crunch, then release it again. I do not think the distribution model is favoring High Performacne machines. I have an i7-975 and an I7-965, both High Performance, very reliable machines that do nothing but crunch WCG 24x7. They have each recieved only about a dozen WUs, since the beginning. I have 96 cores, working on WCG and I have only recieved 296 WUs for a total of 55 days and a silver badge. So I have no idea how some else has got saphire already. At least HFCC is now getting a lot of work from me with now over 15 years worth of work contributed while I wait for DDDT2. If it ever breaks free I could get Saphire in 8 days! ![]() ![]() ![]() |
||
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 2986 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
It's NOT a question of high powered machines, it's WHEN those machines connect to the Grid. If they time it for when there's work available, then they'll grab some WU's - if not, then they won't. What's more, if they time it to connect when "A" types are around, those WU's will last for some 50-100 hrs, whilst if they connect whilst there's "C" types on the go, then those WU's are extremely small, tiddly ones, quite often less than 1 hr.
----------------------------------------Please, please, please remember, we've only crunched an extremely small number of WU's - and thus, there's tons more to come. Time and time again, also remember what the word "Intermittent" means - i.e. there'll be periods when, not even the WCG techs can do anything to generate more WU's, as we're waiting on the scientists to validate/verify everything. ![]() |
||
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3715 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Please, please, please remember, we've only crunched an extremely small number of WU's - and thus, there's tons more to come. Yep, right now 147,978 done, about 44 million more to do. Remember that uplinger's now famous reference chart gives numbers before applying the necessary dual redundancy. That leaves hope, no? ![]() |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
[...]Remember that uplinger's now famous reference chart gives numbers before applying the necessary dual redundancy. I would be even happier, if uplinger gives us a weather forcast for the current week...!!! ![]() ![]() |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I do not think the distribution model is favoring High Performacne machines. I have an i7-975 and an I7-965, both High Performance, very reliable machines that do nothing but crunch WCG 24x7. They have each recieved only about a dozen WUs, since the beginning. What has happened is that people connected to DDDT2, like you, got a reasonable number of WU for the supply that was available.I have 96 cores, working on WCG and I have only recieved 296 WUs for a total of 55 days and a silver badge. So I have no idea how some else has got saphire already. At the same time, there were people who happened on the WU release and upped their buffers at the time. With the release spanning 2 days, they were able to pull the max of 80 WU x 2 days x number of cores and since the WU release had a finite number of WU, the distribution got skewed. Add to that was that the WU release was metered out slowly, so people connected to multiple projects pulled down a WU from another project to fill their normal buffer because there was often no DDDT2 WU when they polled the server. The only people who got a large number of WU were people who had only selected DDDT2 and they were either just plain lucky or they have no life and were waiting to pounce. |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I've heard the pleas for patience. But if this is to be thought of as an intermittent project, and not a project released to be crunched, then the admins should put it in either Beta or Intermittent status. The science is, in the end, what we're all hoping to contribute to. But WCG constructed the 'Ladder of Badges' in part to keep folks involved in crunching by giving them something tangible for their efforts. So, in a community of 'all for the science', it is disconcerting to see what appears to be a distribution model that favors certain high performance machines, rather than distributing the 'science' to all participants. I like the 'beta or intermittent' characterization for what is happening with DDDT-2. I can get behind that knowing that the beta process does have requirements for things like WU return times. The sooner the beta is back, the sooner people can look at the results and make needed corrections, etc. So, pull DDDT-2 back and call it what it is, a beta. When the bugs are worked out and it is ready for EVERYONE to crunch, then release it again. I do not think the distribution model is favoring High Performacne machines. I have an i7-975 and an I7-965, both High Performance, very reliable machines that do nothing but crunch WCG 24x7. They have each recieved only about a dozen WUs, since the beginning. I have 96 cores, working on WCG and I have only recieved 296 WUs for a total of 55 days and a silver badge. So I have no idea how some else has got saphire already. At least HFCC is now getting a lot of work from me with now over 15 years worth of work contributed while I wait for DDDT2. If it ever breaks free I could get Saphire in 8 days! Favoring high performance machines????? 1. 20 day return time, issued as repair with 8 days, reliable machines only, error out a HPF, disqualified 2. machines with 2 day average turn or less, if you consistantly take 50 hrs, not fast enough, disqualified 3. memory requirement 4X actual used, "for faster turnaround" No I don't see any favoring ![]() |
||
|
sk..
Master Cruncher http://s17.rimg.info/ccb5d62bd3e856cc0d1df9b0ee2f7f6a.gif Joined: Mar 22, 2007 Post Count: 2324 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Favouring exists, and I think it should at this stage; to speed up the project, to the point where everyone can contribute, but the project has been undermined by badge hunters that set their cache to 10days and then fail to deliver. This is a serious problem and the way it has been handled, or not, is disturbing.
During the "55days" I had about 30cores attached 24/7 and was only running one other project. I got less than 200 tasks all in. I also had about the same number of cores running part time. A lot less than the 5000+ tasks I could have got by massaging the buffers. I dont like the way grabbing these tasks effects other users with only one or two cores, or the way it slows the research down, so I wont be contributing to the badge grab culture WRT this project - I would prefer such users were not deterred to the point that they leave. The same applies to the Betas; I will on occasion try to get a few work units, but I will not set my cache to many days. Doing so deters participation and is an overall negative for WCG. Hopefully the next time tasks are to be released, I will see an announcement for new tasks; that way I will set my cache to somewhere between 1 and 2 days and get plenty of work units - and hopefully prevent some from filling their cache with 10days worth of Work Units. My tasks will of course actually return well within that time, rather than fail to complete in time, which has been the case with other users trying to grab too many tasks. This project has been slowed down enough as is. When tasks fail to return in time, as has been happening since the outset, they have to be re-issued. This delays the next stage unnecessarily. Be reasonable, you will get a Blue badge in due course. |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
It´s raining ts02 pe /se / Only a short shower
----------------------------------------Simply by a Quad 9450 and a Netbook Celeron Greetings Horst [Edit 1 times, last edit by Former Member at Apr 8, 2010 8:23:30 AM] |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The Deluge begins ..??
----------------------------------------I hope so I just want to crunch and achieve my goals.. Heheheh very sneaky putting the project on line just as Europe is waking and the Americas are sleeping. ![]() ![]() [Edit 2 times, last edit by Former Member at Apr 8, 2010 6:19:27 AM] |
||
|
|
![]() |