| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: Website Support
Thread: curious about priority of ending projects |
| No member browsing this thread |
|
Thread Status: Locked Total posts in this thread: 26
|
|
| Author |
|
|
nasher
Veteran Cruncher USA Joined: Dec 2, 2005 Post Count: 1423 Status: Offline Project Badges: 
|
task priority has always made me wonder..... but now its getting even more confusing to me.
----------------------------------------according to Sekerob's priority sheets. HCC is down to ~ 56 days at medium priority HCMD2 is down to ~ 64 days at medium priority HFCC is down to ~120 days at low priority I understand HFCC being down at low due to project 7 ending and project 8 not being ready yet. I kinda understand them trying to hold HCC till HCC-2 is close to being ready I just do not understand why HCMD2 is down to medium priority. I love being able to see projects completed but I am curious why these projects are all slowing down at the end. so i guess my question is ... is there a reason we are slowing projects at the end like the scientists want it that way or is there just something that my lack of common sense preventing me from seeing  |
||
|
|
astrolabe.
Senior Cruncher Joined: May 9, 2011 Post Count: 496 Status: Offline |
I am nearly certain the decision was made arbitrarily with no relationship either to WU volume or to server capacity at either end
 |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 3010 Status: Offline Project Badges: 
|
I think HCMD2 is slowly being ramped back up - after it's long (too long) outage earlier on in the year. Hopefully, if it hasn't already been, it'll be 'upped' to normal priority very soon.
---------------------------------------- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
On HCMD2, the only thing seemingly missing is the official declaration. At over 120,000 results validated yesterday and 110,000 it's pretty normal to me. But then, do you know what pipe width they have sending the results directly to Paris instead of Lyon?
As to astrolabe.'s comment, like it [with grin]... Conspiriitis is rampant on the NA continent ;>) |
||
|
|
nasher
Veteran Cruncher USA Joined: Dec 2, 2005 Post Count: 1423 Status: Offline Project Badges:
|
what conspiracy... i didn't mention anything like that... hides....
----------------------------------------i am just interested in knowing more of the how and why and what-for for the projects running, new projects, ending projects and whatever else i can learn about WCG and the projects... I fully understand why they don't mention the new projects before they start... I wish they did but... Since I know I will not know of new projects its the fact I see what I would consider strange behavior at the end of projects... this is again looking at SekeRob's wonderful charts HFCC yep they are in a end of project 7 waiting for project 8 so I understand we are waiting for the product to crunch HCC from my understanding again its probably keeping work for the people who only want to crunch cancer projects till HCC2 is ready or nearly ready so if i am correct I understand I guess. I seriously doubt if FLU1 and CEP1 should be on his Active/Intermittent Research Dashboard but I'm not making the chart so its just a comment. I am actively watching the end of HCMD2 and wondering if it is going to get a boost or a slowdown or whatever... I know with the restarting of c4cw its going to be interesting watching the changes over the next few days as well. I understand completely why dddt2 is in its own world the scientists have lots of mid term work I just wish they could load us up with more type A work units but I am guessing its about storage of the results for type A while making type B and C's. I am also curious why DSFL is a medium instead of a normal... but i guess its probably just the scientists are going AGGGGGGGGHHHH as the results from our crunching machine started hitting them that WCG gave them a throttle back.... but again im just a cruncher working with 5 cores doing what i can would love to know more of the insides of WCG though |
||
|
|
Hypernova
Master Cruncher Audaces Fortuna Juvat ! Vaud - Switzerland Joined: Dec 16, 2008 Post Count: 1908 Status: Offline Project Badges:
|
For me it is strange to have the feeling that the WCG grid is just too powerful for some projects, and that scientists are flooded by results and have to throttle back. It gives me a feeling that the scientists have underestimated the crunching power available, and by that have not made the best use possible.
----------------------------------------But on the other side we have CEP2 that is weeping that their project progresses too slowly, means not enough crunching power here. In that specific case they have defined technical requirements that only match a small set of the available machines out there. WCG knows what the average machine type is and how capable it is. Wasn't it possible for the scientists to tailor the WU's adequately ? I know that everybody is working very hard (thank's a lot to tech  ) to make the projects available and running on the grid, and also that it is very difficult to design the perfect WU to mach the incredible diversity of hardware and OS's out there, but maybe the scientists are not well enough informed about how to best use the existing crunching power. |
||
|
|
pcwr
Ace Cruncher England Joined: Sep 17, 2005 Post Count: 10903 Status: Offline Project Badges:
|
From what I understand, there is always two sides:
----------------------------------------The WCG crunchers, who want as much WUs to process asap, against the scientists, who can only release WUs for crunching and analyse the results with the manpower that they have. The scientists can only release new WUs once they have analysed the results of previous WUs. The analysed WUs often give results that the scientists were not expecting, resulting in rethinking of the science and more WUs to process to check. Which often means that the WCG crunchers need to wait for the WUs to be produced and released. <Gets coat and leaves to crunch some more WUs.> Patrick  [Edit 1 times, last edit by pcwr at Dec 8, 2011 10:01:49 AM] |
||
|
|
Hypernova
Master Cruncher Audaces Fortuna Juvat ! Vaud - Switzerland Joined: Dec 16, 2008 Post Count: 1908 Status: Offline Project Badges:
|
From what I understand, there is always two sides: The WCG crunches, who want as much WUs to process asap, against the scientists, who can only release WUs for crunching and analyse the results with the manpower that they have. The scientists can only release new WUs once they have analysed the results of previous WUs. The analysed WUs often give results that the scientists were not expecting, resulting in rethinking of the science and more WUs to process to check. Which often means that the WCG crunches need to wait for the WUs to be produced and released. <Gets coat and leaves to crunch some more WUs.> Patrick Patrick I agree with you. I have missed that important point. The perfect extreme example is DDDT2 where it takes so much time to analyze results and come back with new WU's to be crunched that the project is behaving as intermittent. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Slowly the reality is sinking in. This is not a manufacturers conveyor belt churning out widgets.
------------------------------------------//-- P.S. nasher, where did you get the message that HCC1 is slowed down for HCC2 launch? It's been slowed down because there are technical issues at the scientists and still running at a pace of 200,000 WUs per day. That library [already twice extended] seems near inexhaustible. 225 million completed, in quorum 2, i.e. ballpark net 105 million or so. What we're waiting for is the early 2012 GPU beta port of HCC1, to run concurrent with CPU. [Edit 1 times, last edit by Former Member at Dec 8, 2011 9:41:55 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
snip
I am also curious why DSFL is a medium instead of a normal... but i guess its probably just the scientists are going AGGGGGGGGHHHH as the results from our crunching machine started hitting them that WCG gave them a throttle back.... but again im just a cruncher working with 5 cores doing what i can If that were only true, but unfortunately following the DSFL forum, there's a 195 error problem with this science that was hoped to get identified with the last beta. The project was slowed down to medium, so that those that don't see the issue can crunch on, whilst the general pool is not seeing very many. Whilst the priority status is noted on the Dashboard [with some my side translation at times from what the techs say], the QuickLook chart has a little comment section, 3rd in the list on Charts-1. It currently says "Medium: Beta expected for hangs & other." I've not updated to say the Beta ran, and then when word comes, will tell what findings were. The answers are there in summary, but as with horses and ponds, cant force the members to read them.  --//-- |
||
|
|
|