| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Outsmart Ebola Together
Thread: Starting OET1 "OM" type workunits |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 126
|
|
| Author |
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7844 Status: Recently Active Project Badges: 
|
Birthday Valid? Is that the name of your device? The name of the device is "Birthday." The status of the each of the work units is valid. I named the device Birthday because it was a present from my middle child for my birthday a number of years ago. I am moving this machine to a different project because it is not efficient witrh OEM as indicated by the length of time it takes to process each work unit and the point totals (not thta I care so much for the points). I have one work unit which took over 36 hours to finish which seems unusually slow even for this machine. Q6600 - Windows Vista. Cheers
Sgt. Joe
----------------------------------------*Minnesota Crunchers* [Edit 1 times, last edit by Sgt.Joe at Apr 7, 2018 12:00:34 PM] |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Recently Active Project Badges:
|
(And TPCBF, I feel your pain! Those stuck WUs are a serious pain in my back end when I have not checked a computer in a while and I see them there. I only have three computers at the house, but two are running without monitors. I only "remote" into them when I think there is a problem. My first indication is always that my points are dropping through the floor.) Well, as of this morning, I have multiple pages of results as "no-reply" or "too late". And I can't even fast track all those those resends that some of my hosts get (a lot of _4 and _5, which I haven't seen many of in the few years that I an crunching on WCG now at all). A huge waste of resources...  OET1_ 0000631_ xMBGP-OM_ rig_ 97197_ 1-- ToshibaLatop Too Late 3/27/18 11:35:09 4/7/18 07:00:00 56.58 / 56.58 39.4 / 0.0 Ralf  |
||
|
|
madmac
Advanced Cruncher England Joined: Dec 4, 2005 Post Count: 104 Status: Offline Project Badges: 
|
I too have aborted a few Wu's and may have to do a few more. Most of mine are taking around 14 hours to do one
---------------------------------------- |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Recently Active Project Badges:
|
I too have aborted a few Wu's and may have to do a few more. Most of mine are taking around 14 hours to do one Well, you can do that if you have just a few hosts with direct access. That's true for me for probably about a third of all hosts. The rest are just getting stuffed up.And to make matters worse, pretty much all hosts across the board now start to pick up WUs from other, not selected WVG projects, including FAH2 with its ridiculous short 24h deadline, which now push in front of those ultra long running OET WUs... Ralf [Edit 1 times, last edit by TPCBF at Apr 8, 2018 11:39:49 PM] |
||
|
|
KerSamson
Master Cruncher Switzerland Joined: Jan 29, 2007 Post Count: 1684 Status: Offline Project Badges:
|
For members struggling with host monitoring, I can only recommend you to install BoincTasks on one machine with monitor and to take a look on all your connected hosts from time to time. It is very effective and it does not take much time (only few seconds/minutes).
----------------------------------------Regarding the WUs aborted by server, it is not a huge waste of resources, since the most of these WUs are still not processed at this time. It "wastes" only network resources for redistributing the WUs to another members (however OET1 WUs are very small) and it costs some CPU time on the distribution hosts. It is finally negligible. The move from "normal" to long WUs is disturbing but it is not causing so much resources wasting. Happy crunching, Yves |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Recently Active Project Badges:
|
Regarding the WUs aborted by server, it is not a huge waste of resources, since the most of these WUs are still not processed at this time. It "wastes" only network resources for redistributing the WUs to another members (however OET1 WUs are very small) and it costs some CPU time on the distribution hosts. It is finally negligible. Those WUs are not aborted by the server, but you might have to abort them manually. Wouldn't be that much of a problem if it happens before they have started, but that isn't always the case. Specially when already started WUs are being pushed back by those 24h deadline FAH2 WUs and then result in a lot of "no reply" or "too late" WUs. THAT is a waste of resources...Ralf |
||
|
|
KerSamson
Master Cruncher Switzerland Joined: Jan 29, 2007 Post Count: 1684 Status: Offline Project Badges:
|
Hi Ralf,
----------------------------------------currently I compute only OET1; for this reason I did not experience any "back pushing" by other projects. Initially, I waited and I noticed that a part of the overdue WUs were aborted by the server. Despite of it, I aborted later manually enough WUs I was aware that my hosts would not achieve within the expected time in order to limit redundant and maybe not honoured work. Since I supervise at least 4 of the 6 hosts using BoincTasks, the situation does not represent an issue on my side, excepted if I am travelling at the same time. Cheers, Yves |
||
|
|
Dayle Diamond
Senior Cruncher Joined: Jan 31, 2013 Post Count: 452 Status: Offline Project Badges: 
|
My average completion time has increased to 5.7 hours on the two rigs (1700x, 1950x) I have crunching OET on Ubuntu at stock speeds.
|
||
|
|
highwind
Advanced Cruncher Joined: Aug 22, 2009 Post Count: 56 Status: Offline Project Badges: 
|
I have three machine:
1. 7700HQ - Average is about 10.5 to 11 hours per task 2. 8700K OC @ 5.2Ghz - Average is about 4.4 hours per task 3. Ryzen OC 1920X @ 4.4Ghz - Average is about 5 hours All of my machines fans are blowing off like crazy and this is how my machines should be used! |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Recently Active Project Badges:
|
Hi Ralf, Well Yves, same here. currently I compute only OET1; for this reason I did not experience any "back pushing" by other projects. Ok, I have selected both OET1 and HST, but WUs for the later are in extremely short supply, so practically, I only crunch OET1. However, I now get more and more inundated with WUs from all other active projects, even though I DO NOT have those selected. And among those "non-subscribed" projects, FAH2 with it's 24h deadline is in particular wrecking havoc and causing no-reply and too late results... Ralf |
||
|
|
|