| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: OpenZika
Thread: Limited Launch Again? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 55
|
|
| Author |
|
|
Jason1478963
Senior Cruncher United States Joined: Sep 18, 2005 Post Count: 295 Status: Offline Project Badges: 
|
 Great project!  [Edit 1 times, last edit by Jason1478963 at May 20, 2016 6:29:18 PM] |
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges:
|
My guess is that we took in all of the current batch, just give them some time to release more WUs. While a few machines have a day worth of work, unfortunately a copy of my machines are running dry too.
---------------------------------------- |
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
We are still building more work. I set it to load 50k work units at a time, apparently I have underestimated the amount of work that would be pulled down. The Loading will adjust, but just for good measure I increased this value to 75k.
Thanks, -Uplinger |
||
|
|
Jason1478963
Senior Cruncher United States Joined: Sep 18, 2005 Post Count: 295 Status: Offline Project Badges:
|
Thanks, Uplinger, getting some now.....crunching on :)
---------------------------------------- |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Offline Project Badges: 
|
We are still building more work. I set it to load 50k work units at a time, apparently I have underestimated the amount of work that would be pulled down. The Loading will adjust, but just for good measure I increased this value to 75k. No wonder the initial batch got sucked up that quickly, as they came in with a far to low estimated runtime, hosts just got a very large number to fill the allotted time window in the BOINC agent, even when they end up running up to and over 10x as long as that initial estimate. Luckily, they are set with a 10 day return date or a lot of them would not be returned in time.This is something that IMHO should have been caught during the beta run, the same goes for the hanging .png file downloads. Ralf |
||
|
|
SekeRob
Master Cruncher Joined: Jan 7, 2013 Post Count: 2741 Status: Offline |
You already had a x-posted answer by seippel on the runtime issue!
|
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Offline Project Badges:
|
You already had a x-posted answer by seippel on the runtime issue! Sounds like you're missing the point here, in the context of the supply of OpenZIKA WUs apparently currently having run dry right now...Ralf |
||
|
|
SekeRob
Master Cruncher Joined: Jan 7, 2013 Post Count: 2741 Status: Offline |
No, you failed to indicate to the readers the reason was explained why it happened, as a consequence NOT avoidable. ;O)
----------------------------------------[Edit 1 times, last edit by SekeRob* at May 19, 2016 6:17:55 PM] |
||
|
|
UBT - JohnR
Cruncher Joined: Apr 30, 2006 Post Count: 35 Status: Offline Project Badges:
|
If people had set a cache of 5 days to try and get TB work they would have downloaded Zika work estimated at 15 minutes actual about 3 hours ( 12 times the estimate) That is 60 days worth of work could have been downloaded.
Later changes in the estimated time is no use after the damage has been done. Then Marist College has downloaded over 50K work units. |
||
|
|
SekeRob
Master Cruncher Joined: Jan 7, 2013 Post Count: 2741 Status: Offline |
No, because there is a per-core cap of maximum 35 or so workunits, a safety net for when estimates are off, or simply to protect against overbuffering in cases of high variability in run times.
----------------------------------------NB this cap is all projects of WCG combined! [Edit 2 times, last edit by SekeRob* at May 19, 2016 8:00:09 PM] |
||
|
|
|