| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Discovering Dengue Drugs - Together - Phase 2 Forum
Thread: New thread for It's raining Dengue, Hallelujah!... |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 2370
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I like linux. Whenever I see that there is some rain I enlarge my caches and start a simple script: while true; do sleep 60; BOINC/boinccmd --project www.worldcommunitygrid.org update; done Oh boy, this will surely unleash the fury from SekeRob  |
||
|
|
nasher
Veteran Cruncher USA Joined: Dec 2, 2005 Post Count: 1423 Status: Offline Project Badges: 
|
yes i am seeing work units now and then making it to my computers... I am mostly concentrated on cep2 and c4cw right now working on upgrading those badges but i also leave my cache open for dddt2 i think i am getting one or so a day right now. hope the rain keeps comming
---------------------------------------- |
||
|
|
Jack007
Master Cruncher CANADA Joined: Feb 25, 2005 Post Count: 1604 Status: Offline Project Badges:
|
wouldn't it be better to focus on Dengue while it's raining?
----------------------------------------Although for CEP2 I am still working on 1 or 2 per box, just because i can't focus solely on it due to the high write times. It's up to you, just saying...  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have 170 In Progress for DDDT2 and I have completed all WUs for CEP2 for which I have well over a year and a half. Each CEP2 WU uses 12+ hours of elapsed time if all 16 jobs are able to complete. I want to give priority to DDDT2 at this time since each WU runs for 1+ to 6+ hours and I have yet to reach one year for DDDT2.
For some reason, I didn't get any time recorded this morning. I did check all systems and found that one had 12 WUs ready to be reported as completed. |
||
|
|
sgoll
Advanced Cruncher Joined: Oct 24, 2006 Post Count: 87 Status: Offline Project Badges:
|
I like linux. Whenever I see that there is some rain I enlarge my caches and start a simple script: while true; do sleep 60; BOINC/boinccmd --project www.worldcommunitygrid.org update; done Oh boy, this will surely unleash the fury from SekeRob Well ... I think I'm on the safe side. The application has a 12 second "redial" limit, one can see this "Communication deferred 00:00:11" when pressing the "update" button in the Boinc manager. I use a "security factor" of 5, my script asks every 60 seconds for updates. Of course I know that this will put some stress on the WCG servers, but .... Yes, I can use any other number to wait for updates, but it would be nice when SekeRob or any other could give a hint. The boinc application also asks every 60 seconds for work ... at least for the first tries. Then the time between the updates seems to get longer and longer ...and the my crunchers will miss the raindrops from DDDT2. So the question is: how it is done right? Stephan  |
||
|
|
My 2 cents worth
Cruncher Joined: Sep 12, 2008 Post Count: 19 Status: Offline |
I have a question on my setup.
----------------------------------------On my Linux box 4 cores only project is DDDT2 and it has over 50 WU's ( I Like that part ) but what it is doing is crunching some for a while and then crunches others for a while and so on. Now I have 27 WU's that are partially done, how can I set it up to just do 4 until they are done then go to the next 4. Thanks for any help on this.   |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Plz share the whole story with us. What is the run state of these tasks as they were alternating with 50 of these in cache and what client version are you running, and what is the CPU% and daily hours the client is/was on? A copy of your message log whilst it is swapping / preempting tasks will tell us more. Something like 50 lines.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I like linux. Whenever I see that there is some rain I enlarge my caches and start a simple script: while true; do sleep 60; BOINC/boinccmd --project www.worldcommunitygrid.org update; done Oh boy, this will surely unleash the fury from SekeRob Well ... I think I'm on the safe side. The application has a 12 second "redial" limit, one can see this "Communication deferred 00:00:11" when pressing the "update" button in the Boinc manager. I use a "security factor" of 5, my script asks every 60 seconds for updates. Of course I know that this will put some stress on the WCG servers, but .... Yes, I can use any other number to wait for updates, but it would be nice when SekeRob or any other could give a hint. The boinc application also asks every 60 seconds for work ... at least for the first tries. Then the time between the updates seems to get longer and longer ...and the my crunchers will miss the raindrops from DDDT2. So the question is: how it is done right? Stephan No Fury, just going to load you up with a guilt complex, if I can: At 60 seconds, take 210,000 active devices and multiply that by 1440 and you know the hits on the server and more importantly, the scheduler could be facing in a day. Some tweaks we do not post, simply because if they become prevalent, WCG will have to buy more hardware. No, not going to advise on what is acceptable, as the BOINC Client is meant to manage work fetching in collaboration with the projects scheduler/feeder, though it is not necessarily faultless in doing that job. Use your good judgment, silently ;D -- From "Saturday Night Live"... SekeRob |
||
|
|
My 2 cents worth
Cruncher Joined: Sep 12, 2008 Post Count: 19 Status: Offline |
Plz share the whole story with us. What is the run state of these tasks as they were alternating with 50 of these in cache and what client version are you running, and what is the CPU% and daily hours the client is/was on? A copy of your message log whilst it is swapping / preempting tasks will tell us more. Something like 50 lines. Well I am running Fedora 14 Linux, Boinc Version 6.10.58 Everything runs great on all 4 cores. It is on all the time 24-7 with no interruptions. The only thing that happens is for whatever reason I get 4 WU's that are not close to date example due on 30 Jan that want to run in High Priority, so they take over, then when they are done 4 different WU's seem to want to start up and work until another 4 WU's decide they want to run in High Priority and it just keeps going like that. So I end up with a slew of WU's that are all different % of being complete and they will get around to finishing at some point I am not worried about that, it just looks like an unorganized mess vs my Mac machine that is set up with the exact same settings and never gets run in high priority mode WU's that I am aware of. This is not a huge issue, everything is working fine, I was just wondering if there is a way to tell it to nock it off with the running in high priority mode. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Well, something in your client is thinking that your cache up and time to deadline makes the scheduler compute that some will have trouble to meet deadline. If before your client was a fast returner ** getting repair jobs, usually with a 4 day deadline and your cache or switch times are too high (cache larger than shortest or nearest deadline), you get served with these alternations, but grant you, the developers are even now battling in 6.12 to stop this senseless trying business (to find the shortest task e.g.)... assuming those 27 partially done tasks are all WCG.
** under 2 days. Then if you up the cache to quel the thurst on that DDDT2 rain, it takes the servers more than 2 days to realize that the client is no longer a fast returner before it stops sending short deadline jobs. |
||
|
|
|