| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: FA@H Single Redundancy Change |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 92
|
|
| Author |
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
Greetings all,
We are about to start the change to single redundancy for FightAIDS@Home. This means over the next week members may notice no work being available if they have FA@H as the only project selected. The advantage of this switch will allow the grid to run more efficiently as we were sending out 2.1 work units for every result. This should go down to about 1.2 by our estimations. This will allow us to process more work units for the researchers by using the same amount of processing power. An added benefit is that members will also have their results validated quicker. Thank you for your patience as we switch the system to single redundancy. (Please note this is the same change that has happened to Discovering Dengue Drugs - Together, and there is quite a bit of information about the change details in this thread. https://secure.worldcommunitygrid.org/forums/wcg/viewthread?thread=21397 -Uplinger |
||
|
|
Rickjb
Veteran Cruncher Australia Joined: Sep 17, 2006 Post Count: 666 Status: Offline Project Badges: 
|
Recently, we've started crunching the longest FAAH WUs that I've seen since I started crunching in late 2006. About 1 week ago, we had the first of these, and they took 4 - 5 x the "normal" time. These were followed by a few days of the previous shorter WUs. Then 1 or 2 days ago, even longer WUs, up to about 7 - 10 x "normal", started to arrive. BOINC seems to base its estimates of crunch-time on recently-crunched WUs, and when there is a sudden increase in crunch time, it fetches too much work into our work queues.
Remembering the multi-day WCG server outage at Christmas 06, I normally run a 3-day work cache, but fortunately recently reduced this to 0.5 days. Even so, the WUs currently in the q for my A64 will take about 80hrs to finish. There will be others out there with far more work. Having this happen when you need to empty the work queues to switch to Single Redundancy seems to be unfortunate timing. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I don't want to distract this thread, so who ever reads about these very long running FA@H tasks and has one, please post the job name here . This allows others to see what particular batch this is. There are several in circulation, long and normal. The long are not prevalent as the average project daily run time reported is still relatively low with 8.85 hours (light blue curve in below graph). So please discuss that topic in there:
----------------------------------------As for the nearing end and reaching the point where the techs will switch over the version, the simplest is to watch this graph. Many more you find when following the link below in my signature. Members who have selected to only run FightAIDS are strongly recommended to go to My Grid > Device Manager, select the device profile associated with your client(s) and check that the "If there is no work available...." box. When FA@H temporarily has no assignments for the interim period, WCG will send other jobs to keep your client busy. It is very much appreciated. thanks
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Just set a device profile to only get FA@H and the "if no work is available" option and got 3 RICE, so presume that the regular FA@H work is depleted and only repair jobs are out there until the Techs start forcing out the 3rd redundancy copy to speed up the clean out.
----------------------------------------ciao PS Oddly, there was no message in the log to say there was no work available for the selected projects.... hmmm. A repeat of the trick gave me 3 DDDT and still no message.... very hmmm.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Jul 31, 2008 10:25:50 AM] |
||
|
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges: 
|
We have another 24-36 hours of regular FAAH work to send. So you should still be able to work for it.
----------------------------------------[Edit 1 times, last edit by knreed at Jul 31, 2008 11:54:26 AM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Just this morning, almost 4 days after I still get fresh work, this morning 8 out of the faah4259 batch.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
petehardy
Senior Cruncher USA Joined: May 4, 2007 Post Count: 318 Status: Offline Project Badges: 
|
Hi Guys,
----------------------------------------I was just wondering what's happening. It looks like experiments 19a and 24 are in resend mode, are we trying to finish these before changing over to single redundancy? Also I'm getting a lot of experiment 20 WUs, do these need to be completed as well? I'm guessing that it's about 60% done. Thanks, Pete  "Patience is a virtue", I can't wait to learn it! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
We are currently helping out the scientists with a priority job to allow them to complete a paper. As soon as this priority work is finished we’ll switch to single redundancy.
|
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges: 
|
Pete, you might also look at these threads for more details
----------------------------------------experiment confusion now really short WU's Long WU Cheers. Jean. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Hi Guys, I was just wondering what's happening. It looks like experiments 19a and 24 are in resend mode, are we trying to finish these before changing over to single redundancy? Also I'm getting a lot of experiment 20 WUs, do these need to be completed as well? I'm guessing that it's about 60% done. Thanks, Pete My take on this is, that there are lots of crunchers who prefer FAAH. Long as we wait on the monsters to come back (the priority batch), why not send out the short and sweet stuff. Run times I've seen yesterday were under 3 hours.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
|