| 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 |
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 3010 Status: Offline Project Badges: 
|
Well, across my 2 machines, so far I've recieved 21 WU's - the first 2 of which, didn't count against my total as one errored out whilst the other was too late (both of these were on my 'slower' machine -I've since had another on that machine and, due to kicking it off as soon as I spotted it, made it back in time).
----------------------------------------The last WU I recieved, was 8 days ago - so, although I'm certainly not complaining (after all, as has been said elsewhere, it's expected that this project slowly gets ramped up), currently it'll take me quite some time to even reach Silver level.  |
||
|
|
DCS1955
Veteran Cruncher USA Joined: May 24, 2016 Post Count: 668 Status: Offline Project Badges:
|
Today my "Reliable" machine which gets resend WU's got 6 -0 WUs, along with 3 HSTB. If this keeps up I will not have squirt in MIB WU's to keep it filled up.. Nice problem to have.
----------------------------------------  |
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges: 
|
No arp1 for 2.5 days.
Mike |
||
|
|
flensr
Cruncher Joined: Oct 31, 2018 Post Count: 25 Status: Offline Project Badges: 
|
Very interesting, I'm getting a lot now. It's possible WCG didn't like my machine working on more than one project even though I set up priorities to give WCG 2-4 cores pretty much full time. Once I suspended the other projects (seti, rosetta, climateprediction) I've gotten about a dozen ARP WUs in the last 2 days.
----------------------------------------I've set up the task preferences to ask for unlimited ARP, and only up to 3 WUs for each of the other WCG projects. I got my first ARP WU with it set to only 2 WUs for each non-ARP project but I ended up with idle CPU time. Upping it to 3 per project has kept the machine working almost full time, right now working on 3 ARP WUs with 2 more in the queue. Feast or famine? Dunno how it really works but dedicating the machine to WCG and putting it right on the edge of running out of work with an idle core or two on occasion seems to have worked.  |
||
|
|
DCS1955
Veteran Cruncher USA Joined: May 24, 2016 Post Count: 668 Status: Offline Project Badges:
|
Very interesting, I'm getting a lot now. It's possible WCG didn't like my machine working on more than one project even though I set up priorities to give WCG 2-4 cores pretty much full time. Once I suspended the other projects (seti, rosetta, climateprediction) I've gotten about a dozen ARP WUs in the last 2 days. I've set up the task preferences to ask for unlimited ARP, and only up to 3 WUs for each of the other WCG projects. I got my first ARP WU with it set to only 2 WUs for each non-ARP project but I ended up with idle CPU time. Upping it to 3 per project has kept the machine working almost full time, right now working on 3 ARP WUs with 2 more in the queue. Feast or famine? Dunno how it really works but dedicating the machine to WCG and putting it right on the edge of running out of work with an idle core or two on occasion seems to have worked. Possiby good appoach. I appeared to have just been lucky in getting recognized . I got close to gold on one slow machine by scheduling requests. Since that quit happening my new faster machine has snagged a lot plus short turn around resends. Since then I get 4-5 and some rare HSTB WU every other day. I keep a 2 day cache of work. It all may be a crap shoot, but I appear to have stumbled on to something. |
||
|
|
sptrog1
Master Cruncher Joined: Dec 12, 2017 Post Count: 1593 Status: Offline Project Badges: 
|
I have been getting an average of one task per day since onset. I run two machines (one dedicated to WCG projects) and return results within 3 days.
There was a time when arp units took forever to finish and the forums seemed to suggest that the immunity project was constipating the cache, so I reduced the number of those tasks allowed and ran some others instead, All that seems to be making a difference. But I still average around 30 hrs. |
||
|
|
DCS1955
Veteran Cruncher USA Joined: May 24, 2016 Post Count: 668 Status: Offline Project Badges:
|
I have been getting an average of one task per day since onset. I run two machines (one dedicated to WCG projects) and return results within 3 days. There was a time when arp units took forever to finish and the forums seemed to suggest that the immunity project was constipating the cache, so I reduced the number of those tasks allowed and ran some others instead, All that seems to be making a difference. But I still average around 30 hrs. I have a old laptop that took up to 40 hrs to complete which got a lot of work when you could schedule work. With random distribution my new desktop which executes ARP in 14 hrs, that is the machine that gets both resends an original work.. There may be some preference to fast processors at least for resends and possibly original work. |
||
|
|
KerSamson
Master Cruncher Switzerland Joined: Jan 29, 2007 Post Count: 1684 Status: Offline Project Badges: 
|
I feel the current situation being a little bit depressing.
----------------------------------------With 60 available threads (52 Linux, 8 Windows 7 respectively 44 AMD, 16 Intel) and a small buffer (0.5 day), until now, I was able to only compute 23 WUs for ARP1. Since I am business traveling more or less full time, I am not able to babysit / micromanage each single host. I do not know what I should do more/better for receiving a little bit more ARP work. Cheers, Yves ---------------------------------------- [Edit 1 times, last edit by KerSamson at Dec 15, 2019 1:42:10 PM] |
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges:
|
Yves
----------------------------------------Firstly, all the projects deserve to be worked on, so download others as well as arp1. Set all machines to download all current projects, to spread the load and not to run out on any machine. mcm1 and mip1 seem to be amost permanently available, so there is no need to set your device profile on each machine for those projects to more than 2 above the number of threads on that machine. Those 2 projects will cause updates whenever a unit has finished so could hit the random release of arp1 units if you are lucky. Boinc Manager should make sure that any arp1 units you get finish before its deadline. However they might not finish before the earlier deadline to classify as a 'reliable' machine. (Plenty about that definition in other postings.) Mike [Edit 1 times, last edit by Mike.Gibson at Dec 15, 2019 5:37:55 PM] |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7849 Status: Offline Project Badges:
|
I do not know what I should do more/better for receiving a little bit more ARP work. What has worked for me is very similar to what Mike Gibson has suggested. Set your combined MCM and MIP number to be either equal or 1 less than the number of threads available on that machine. Set at ARP to at least one or maybe 2 if you are adventurous. If you set the number of MCM and MIP to one less than the number of threads you have you may end up with one thread idle for a bit, but your machine will query quite often for the remaining slot to be filled with an ARP unit. Once the ARP unit is in place, you will not have any idle threads for the duration of the length of time the ARP unit runs. If you have the combined number of MCM and MIP equal to the number of threads, you will still query quite often to fill the extra slot with an ARP unit and you will not have any threads idle. I run 70+ threads and average a couple of ARP units per day using option 2. It requires no babysitting at all. Hope this helps.
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
|