| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Outsmart Ebola Together
Thread: Outsmart Ebola Together - v7.19 - Released |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 53
|
|
| Author |
|
|
vfrey
Cruncher Joined: Nov 16, 2004 Post Count: 11 Status: Offline Project Badges: 
|
On my observation there are 8 checkpoints throughout one job regardless of its length. Checkpoint occurs every 12,5 percent. thanks, I was wondering if these WUs have checkpointing at all. So the first WU I have will checkpoint every two hours... |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Knew had seen it somewhere, a tech too speaking of 7 checkpoints at exact 12.5% intervals, so guess the 8th is hidden inside the wrap up.... after all 100 divided by 8 is what it is, 12.5, without mathurbation, so opened up stdoutdae.txt and filtered out...
12-Feb-2015 10:52:58 [World Community Grid] Starting task OET1_0000334_xSDGP-F_rig_9766_1 12-Feb-2015 10:52:58 [World Community Grid] [cpu_sched] Starting task OET1_0000334_xSDGP-F_rig_9766_1 using oet1 version 719 in slot 7 12-Feb-2015 11:30:33 [World Community Grid] [checkpoint] result OET1_0000334_xSDGP-F_rig_9766_1 checkpointed 12-Feb-2015 12:06:08 [World Community Grid] [checkpoint] result OET1_0000334_xSDGP-F_rig_9766_1 checkpointed 12-Feb-2015 12:41:10 [World Community Grid] [checkpoint] result OET1_0000334_xSDGP-F_rig_9766_1 checkpointed 12-Feb-2015 13:15:55 [World Community Grid] [checkpoint] result OET1_0000334_xSDGP-F_rig_9766_1 checkpointed 12-Feb-2015 13:54:03 [World Community Grid] [checkpoint] result OET1_0000334_xSDGP-F_rig_9766_1 checkpointed 12-Feb-2015 14:31:43 [World Community Grid] [cpu_sched] Preempting OET1_0000334_xSDGP-F_rig_9766_1 (left in memory) 12-Feb-2015 14:31:53 [World Community Grid] [cpu_sched] Resuming OET1_0000334_xSDGP-F_rig_9766_1 12-Feb-2015 14:35:04 [World Community Grid] [checkpoint] result OET1_0000334_xSDGP-F_rig_9766_1 checkpointed 12-Feb-2015 15:13:33 [World Community Grid] [checkpoint] result OET1_0000334_xSDGP-F_rig_9766_1 checkpointed 12-Feb-2015 15:49:15 [World Community Grid] [checkpoint] result OET1_0000334_xSDGP-F_rig_9766_1 checkpointed 12-Feb-2015 15:49:19 [World Community Grid] Computation for task OET1_0000334_xSDGP-F_rig_9766_1 finished 12-Feb-2015 15:49:21 [World Community Grid] Started upload of OET1_0000334_xSDGP-F_rig_9766_1_0 12-Feb-2015 15:49:26 [World Community Grid] Finished upload of OET1_0000334_xSDGP-F_rig_9766_1_0 12-Feb-2015 15:49:31 [World Community Grid] [sched_op] handle_scheduler_reply(): got ack for task OET1_0000334_xSDGP-F_rig_9766_1 Conchita would probably tell it's W...  Presently got a specimen running of batch 333 which after 11:30 hours CPU time shows 31.7%, 2 checkpoints. |
||
|
|
mito7
Advanced Cruncher Slovakia Joined: Oct 12, 2008 Post Count: 58 Status: Offline Project Badges: 
|
Of course. My mistake. It is actually 8 segments (each 12.5%) which means 7 checkpoints. 8th "checkpoint" is end of work unit.
---------------------------------------- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
After seeing this...
OET1_ 0000469_ xEBGP-FA_ rig_ 3019_ 0-- Valid 2/13/15 11:09:50 2/13/15 18:36:07 1.68 / 1.69 173.5 / 173.5 OET1_ 0000469_ xEBGP-FA_ rig_ 2973_ 0-- Valid 2/13/15 11:09:50 2/13/15 18:04:55 1.15 / 1.16 125.7 / 125.7 OET1_ 0000469_ xEBGP-FA_ rig_ 3017_ 0-- Valid 2/13/15 11:09:50 2/13/15 17:59:22 0.82 / 0.82 89.6 / 89.6 OET1_ 0000345_ xZAGP-S_ rig_ 10946_ 0-- Valid 2/13/15 00:09:25 2/13/15 09:25:58 2.31 / 2.31 284.1 / 284.1 OET1_ 0000345_ xZAGP-S_ rig_ 4025_ 0-- Valid 2/13/15 00:15:57 2/13/15 08:55:28 1.39 / 1.39 155.6 / 155.6 OET1_ 0000345_ xZAGP-S_ rig_ 3020_ 0-- Valid 2/13/15 00:13:46 2/13/15 04:34:44 1.21 / 1.21 127.4 / 127.4 OET1_ 0000345_ xZAGP-S_ rig_ 5214_ 1-- Valid 2/13/15 00:20:13 2/13/15 04:24:50 0.65 / 0.65 68.8 / 43.0 OET1_ 0000345_ xZAGP-S_ rig_ 10996_ 0-- Valid 2/13/15 00:09:25 2/13/15 03:45:41 0.90 / 0.91 96.2 / 96.2 OET1_ 0000345_ xZAGP-S_ rig_ 1919_ 0-- Valid 2/13/15 00:11:36 2/13/15 03:40:28 0.52 / 0.52 55.6 / 55.6 OET1_ 0000345_ xZAGP-S_ rig_ 9545_ 0-- Valid 2/13/15 00:05:06 2/13/15 03:21:51 0.81 / 0.81 88.0 / 88.0 OET1_ 0000345_ xZAGP-S_ rig_ 11446_ 0-- Valid 2/12/15 23:49:25 2/13/15 03:07:16 2.17 / 2.17 276.3 / 276.3 OET1_ 0000345_ xZAGP-S_ rig_ 8023_ 0-- Valid 2/13/15 00:02:54 2/13/15 02:58:53 0.60 / 0.60 77.5 / 77.5 OET1_ 0000345_ xZAGP-S_ rig_ 6760_ 0-- Valid 2/13/15 00:00:48 2/13/15 02:51:10 0.62 / 0.62 79.2 / 48.7 OET1_ 0000345_ xZAGP-S_ rig_ 6769_ 0-- Valid 2/13/15 00:00:47 2/13/15 02:33:02 0.75 / 0.75 99.7 / 97.6 OET1_ 0000345_ xZAGP-S_ rig_ 3165_ 0-- Valid 2/12/15 23:54:21 2/13/15 02:22:42 1.22 / 1.22 175.8 / 175.8 Slipped into a spontaneous chanting "Linus, Linus, Linus, Linus ..." (Yes 100+ credits per hour... AD VINA is fast on Linux, but not that fast. :) |
||
|
|
3rkko
Advanced Cruncher Finland Joined: Aug 2, 2008 Post Count: 105 Status: Offline Project Badges:
|
No matter how fast the app is on Linux, the vast majority of users are on Windows. So instead of celebrating the Linux-speed, we should be wondering why the technicians haven't made this as fast on Windows?
---------------------------------------- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
We have been wondering and discussing this for as long as can be remembered at the WCG forums. It's a case of Windows simply not having the matching performing compilers/libraries that exist for Linux. The techs always seek the best [choices of hundreds], and do take great effort to get same throughput, but have to take into account that results for all platforms must be closely matching, if not perfectly. So it is what it is, for some sciences the apps perform/performed often 60% better under *nux (and on OS-X).
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
SekeRob:
What were your runtimes like on the 333 series? Those are the long runners on my Win7-64 boxes. |
||
|
|
KWSN-A Shrubbery
Senior Cruncher Joined: Jan 8, 2006 Post Count: 476 Status: Offline Project Badges: 
|
After seeing this... Slipped into a spontaneous chanting "Linus, Linus, Linus, Linus ..." (Yes 100+ credits per hour... AD VINA is fast on Linux, but not that fast. :) Looking at my BOINCTasks graph is amazing. One machine has gone from 7800 RAC to 12,000 in three days and still climbing. Meanwhile, the lone windows machine is flatlined.  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
SekeRob: What were your runtimes like on the 333 series? Those are the long runners on my Win7-64 boxes. Regrettable none left on the RS pages to look at [if they came through on here]. What is seen of 333 is, runtimes were quite spread too, so can't be sure any few would be fair to use as reference. AD VINA is just 60+% faster on Linux [HCC was too]. |
||
|
|
deltavee
Ace Cruncher Texas Hill Country Joined: Nov 17, 2004 Post Count: 4894 Status: Offline Project Badges: 
|
After seeing this... Slipped into a spontaneous chanting "Linus, Linus, Linus, Linus ..." (Yes 100+ credits per hour... AD VINA is fast on Linux, but not that fast. :) Looking at my BOINCTasks graph is amazing. One machine has gone from 7800 RAC to 12,000 in three days and still climbing. Meanwhile, the lone windows machine is flatlined. I'm not saying that Linux isn't faster than Windows, but I'm getting 100+ credits per hour for some WUs on my Win64 machines also. OET1_ 0000344_ xZAGP-FW_ rig_ 6762_ 0-- xxxxxxxxx Valid 2/12/15 21:26:33 2/14/15 02:08:58 0.42 / 0.42 69.7 / 69.7 [Edit 1 times, last edit by deltavee at Feb 14, 2015 12:35:25 PM] |
||
|
|
|