| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Active Research
Forum: Africa Rainfall Project
Thread: Work Available |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 3595
|
|
| Author |
|
|
DCS1955
Veteran Cruncher USA Joined: May 24, 2016 Post Count: 668 Status: Offline Project Badges: 
|
3 resends one with a 4 day dead line, other two have a 3 day .
----------------------------------------  |
||
|
|
hchc
Veteran Cruncher USA Joined: Aug 15, 2006 Post Count: 865 Status: Offline Project Badges: 
|
Observation: I started making ARP1 work units a priority. In other words, as soon as one hits the queue, I suspend other work (e.g. MIP1) so that the ARP1 kicks off. This way it gets returned in <1 day. Ever since I made this change, I no longer get the "Tasks are committed to other platforms" message. I think this means my devices are considered "Reliable" (returns tasks within the 2.5 day window unless it's been changed to 3 or 3.5 days). I'm not seeing any increase in work units because of this, but at least those devices are now eligible for the occasional repair work unit. (I've gotten only one so far.)
----------------------------------------
[Edit 2 times, last edit by hchc at Dec 12, 2019 7:41:40 AM] |
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges: 
|
dcs1955
Please take a closer look. Resends should all be 3 days 12 hours. Mike |
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges:
|
hchc
I suspect that they have tweaked the message because I stopped getting the other platform message a few days ago. Uplinger did say say that the definition of reliable has been changed. Mike |
||
|
|
JEklund2
Advanced Cruncher Finland Joined: Aug 10, 2006 Post Count: 119 Status: Offline Project Badges:
|
BTW, is there any estimation or even statistics how many WU's there are released daily? I have tried to catch any with Boinc on Linux or Windows but haven't had any luck to get even any WU's in few days
----------------------------------------  |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7849 Status: Offline Project Badges: 
|
According to Africa Rainfall Project Project Statistics page the average per day being returned and validated is 1729.11. So I would surmise that the number being released on a per day basis is right around that number as of Dec. 12,2019.
----------------------------------------Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
hchc
Veteran Cruncher USA Joined: Aug 15, 2006 Post Count: 865 Status: Offline Project Badges:
|
JEklund said:
----------------------------------------BTW, is there any estimation or even statistics how many WU's there are released daily? I have tried to catch any with Boinc on Linux or Windows but haven't had any luck to get even any WU's in few days uplinger said that 21 work unit pairs get released every half hour, so that's 42 pairs per hour = 1008 pairs per day = 2016. I think this is the current maximum. They said the goal would be over 30,000 pairs (60,000 total results) in the wild at once, but I don't know what needs to happen before they're comfortable going full speed.
|
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges:
|
Sgt.Joe
Your figures indicate about 72 per hour. Uplinger said recently that there were 21 being released every half hour. On the assumption that he was lumping together the -0 & -1, that would equate to 84 per hour which is close enough to your 72. Mike |
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges:
|
hchc
Your posting was while I was doing mine and saying the same thing. As to your second paragraph, I am guessing that the scientists will be testing out how each 48 hour period we calculate compares with what actually happened in July 2018. Then making some adjustments to the program and hopefully get a greater correlation. I would expect the work to step up as they get more accurate predictions. There is no point going full blast until results closely match what actually happened. On the basis of 30K pairs and 1 year of data, the project would only last 6 months. Hopefully they will go beyond the first 18 months and start using us to predict the next year. Mike |
||
|
|
kinski
Advanced Cruncher Joined: Nov 25, 2006 Post Count: 104 Status: Offline Project Badges:
|
It's a slow trickle, this. I get max 1WU per day..
---------------------------------------- |
||
|
|
|