| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: Have very slow WU # FAHV_x3VQ7_IN_LEDGFa_rig_0220728_0005 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 234
|
|
| Author |
|
|
littlepeaks
Veteran Cruncher USA Joined: Apr 28, 2007 Post Count: 748 Status: Offline Project Badges: 
|
I am also running all the reruns, and have gotten to the point where I need to "micromanage" what's running to make cutoff times.
----------------------------------------As far as I'm concerned, these WUs are much more interesting than the "plain vanilla" WUs we usually get, and I am more than happy to run them. So far, I've had no errors since the "-131 error" has been fixed. Final comment; "Bring 'em on".  lp [Edit 1 times, last edit by littlepeaks at Oct 2, 2014 11:17:24 PM] |
||
|
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
Just aborted a batch that were running on 8 DAYS cpu time and had only reached around 50%. Aborted the lot of them since they would have only errored out eventually.
----------------------------------------We desperately need the promised flag for no repair units.  Distributed computing volunteer since September 27, 2000 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Not seen them on the tab for more than a week now. The great thing for the points hungry, because the projected runtimes are ballooned, but real time is about 6-6.5 or so, the servers were granting up to 42/hr. Now slowly tailing off, latest valid got 33/hr. Considering 6.5/hour was normal, like getting 5-7 times more than normal.
 Converse, the one race horse i have gets about 12/hr under linux where 35-40 is normal for fahv. But for 1 core running a cep2 test series, the rest of the cpu is given a rest from the madness  A repair selector -and- a fahv-faah selector to go, and giving a bonus to motivate repair participation + a badge for time contributed fixing (oflmbo). Remember how quickly wcg technicians were able to push out the cep2 1-16 + unlimited profile picker and the bandwidth override function? Took like 12 hours. Goes to show there's just good old foot square unwillingness, the 'wcg likes to control what you compute as much as possible'. Positing that it confuses members is an insult to the members intelligence. We have been potty trained since long.  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Potty trained, and well able to operate our nodes to be most productive, something, again, wcg is not able, assign heavy work to powerhouses and light to slower and or intermittent hosts, not to speak of the optimal for linux, such as fahv is. As in hacking, blunt application of brute force computing instead of seeking efficiency.
Good night to you all.  |
||
|
|
pramo
Veteran Cruncher USA Joined: Dec 14, 2005 Post Count: 707 Status: Offline Project Badges: 
|
Orangey ones come down pretty quick on the time to complete.
----------------------------------------Long running green ones, usually there are 6-12 running depending on when I look (ignore that top one, its wuprop@home). I just let them run and let Boinc sort em out. Boxes are 24/7 and I don't have any Too Lates (I say with fingers crossed), but there are many that finish 8 to 10 hours before deadline. Those are fun to watch  Unfortunately I can't speak as to whether they'll come up valid or error out and make another resend but Trying my best to get the cr@p out of the way     |
||
|
|
deltavee
Ace Cruncher Texas Hill Country Joined: Nov 17, 2004 Post Count: 4891 Status: Offline Project Badges: 
|
I just let them run and let Boinc sort em out. Boxes are 24/7 and I don't have any Too Lates (I say with fingers crossed), but there are many that finish 8 to 10 hours before deadline. Those are fun to watch Unfortunately I can't speak as to whether they'll come up valid or error out and make another resend but Trying my best to get the cr@p out of the way pramo, I'm having fun watching these long ones too. This one did not error out after two weeks. FAHV_ x3VQ8_ B_ IN_ LEDGFa_ rig_ 0221725_ 0034_ 0-- 716 Pending Verification 9/5/14 07:52:08 9/19/14 12:46:01 180.80 227.7 / 0.0 Like you said, I'm just trying to stay out of the way and let WCG do its thing. All of these long ones will eventually run and it's interesting being a part of it. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Another feature request can always be bumped to the fore, and lacks any originality because it's done, was done since many years: Just like rosetta has the option to pick how many hours, create a profile setting that has the selection short/medium/long. Then, those enjoying the watch for the exponential chance of failing and the degressivity in points per hour can have their joy extended
|
||
|
|
pramo
Veteran Cruncher USA Joined: Dec 14, 2005 Post Count: 707 Status: Offline Project Badges:
|
Another feature request can always be bumped to the fore, and lacks any originality because it's done, was done since many years: Just like rosetta has the option to pick how many hours, create a profile setting that has the selection short/medium/long. Then, those enjoying the watch for the exponential chance of failing and the degressivity in points per hour can have their joy extended I was trying not to think about the hit my points and gFlops are taking  . In fact, I removed a line or two about that from my earlier post.  I'll just say watching the clock tick down is the only fun part deltavee and I have to look forwared to in this mess |
||
|
|
OldChap
Veteran Cruncher UK Joined: Jun 5, 2009 Post Count: 978 Status: Offline Project Badges:
|
I have a rig that I try to make exclusively FAHV. I too saw really poor points when these ran beyond the limit and were returning sixty odd points for a day or maybe two of work
----------------------------------------BUT Now I am running wu's that complete in around half an hour, plus a few long re-sends, and the points are astronomically high. I have a low of 25,000 wcg points, an expectation of 175,000, and a high of 1,982,000 points in one day from this rig. You may yet see your points haul recover from these long ones while your rig tries to re-calibrate to running short wu's so perhaps something to look forward to after all  |
||
|
|
pramo
Veteran Cruncher USA Joined: Dec 14, 2005 Post Count: 707 Status: Offline Project Badges:
|
I have a rig that I try to make exclusively FAHV. I too saw really poor points when these ran beyond the limit and were returning sixty odd points for a day or maybe two of work BUT Now I am running wu's that complete in around half an hour, plus a few long re-sends, and the points are astronomically high. I have a low of 25,000 wcg points, an expectation of 175,000, and a high of 1,982,000 points in one day from this rig. You may yet see your points haul recover from these long ones while your rig tries to re-calibrate to running short wu's so perhaps something to look forward to after all TBH- when i look at it over the long haul, things that i didn't know i would get credit for seems to be trickling in. up-thread some of my concerns were addressed for some of the errors. At that point (other concerns not necessarily addressed directly) i just said I'm not going to worry about it and will crunch what i get. different for everyone and I'm fortunate to have 99.999% connection and several cores (for the moment). i set a .2 day cache (and have been burned maybe 2-3 times in the last several years. when i ran short i found other things to do short term (hi Poem!) i get a wu that runs for 3 hours and gets a 220 points, another runs for 60 hours and gets 68 points...frustrating maybe but i just let it ride.yeah I care some about the time and points and stuff but bottom line I just tell myself in some way it advances the science. Side note to that- One day maybe I'll tell the story about how, running only on a 386/16 I got pissed at long running work units on African Climate@home and Genome Comparison and aborted them. and by doing so, I lost my full boat and fark me. Sorry for the last bit, If you need to ban my posts that's fine but please don't ban my boxes. All that said, by sticking with what i was given and looking forward- i don't know if it's the shorter units coming out, the old long ones getting covered and time/points awarded- but things seem to be looking up over the last day or so  I'm just hanging in  |
||
|
|
|