Welcome to MilkyWay@home

Posts by EdwardPF

1) Message boards : News : New N-Body Release (Message 57180)
Posted 6 Feb 2013 by EdwardPF
Post:
by the way ... It sounds completely crazy BUT ...

While the WU was running my computer clock was 1 hr behind real time (switch time zones??) ... after the WU finished ... my computer time is back to normal ....


Now that is strange!!

Ed F
2) Message boards : News : New N-Body (Message 57175)
Posted 5 Feb 2013 by EdwardPF
Post:
The WU finished with status "success".

Wall time: 379,018 sec
CPU time: 1,315,527 sec CT/WT = 3.471 (I was using 4 CPU's)
credit: 3004. 9

If I was running the standard MW WU's this much CPU time would have yielded about 33000 Cobblestones (o.k. maybe 31000 would be closer ... ??)

But the thing worked!!

Ed F
3) Message boards : News : New N-Body Release (Message 57171)
Posted 5 Feb 2013 by EdwardPF
Post:
My first 1.06 finished last night successfully ... however I don't think the cobblestone Wh...s will be happy with the credit! (Workunit 303830201)

Wall time 379,018 sec (4d23h12m25s), CPU time: 1,315,527 sec, credit: 3,004.90 :-(

if this had been the "de_separation_15_sSgr_1" kinda' runs the credit would have been about 33,000 cobs.

Perhaps a tweak in the code is in order :-)

Ed F
4) Message boards : News : Nbody 1.06 (Message 57154)
Posted 3 Feb 2013 by EdwardPF
Post:
My 2 cents:

The new 1.06 WU are better behaved then the previous release in regards to EST time but still need some refining ;-)

I'm not sure how they can produce a good number for remaining time since the nbody WU's are multi-threaded ... I.E. it can run on one or all of your CPU's at the same time - in spite of what your BOINC params say.

My current nbody is running on 4 CPUs even though I have BOINC set to 12.5% of my ("8") CPU's. its orig est time was 59hrs and has run for 64 hrs with 36.75 hrs remaining ... I'm sure the "boys back at the office" are working on this.

Ed F
5) Message boards : Number crunching : How accurate are Time Remaining estimates? (Message 57149)
Posted 3 Feb 2013 by EdwardPF
Post:
If you are running

de_separation_15_sSgr_1
de_separation_16_sSgr_1
ps_separation_15_sSgr_1
ps_separation_16_sSgr_1


type WUs my experience is that the remaining time was WAY off (250Hrs est for a 3.5 hour run) . but as I processes more and more the EST time homed in to 3:15:45 and hovered there ... a GOOD Est.

However ...

If you are running nbody 1.06 My runs say it's a crap shoot. my current run started with an EST time of 59 Hrs ... it has been running for 51.5 (wall time) hrs on 4 CPU's, (the WU is Multi threaded and will run on 4 cpu's at once if it can), it has accumulated 176.5 CPU hours and has an EST remaining time of 44 Hrs.

I ASSUME that will be about 44 wall time hours (it's reporting 51% + done) and another 175 CPU hours ...

But that's just a guess ...for now ...

Ed F
6) Message boards : News : New N-Body (Message 57135)
Posted 1 Feb 2013 by EdwardPF
Post:
With the WU 18.9% done and 16.5 hours into it ... it has consumed 59 hours of cpu time to date. this is a ratio of about 3.5cpu hrs to 1 wall hour. Some of this is due to computer usage no doubt but the remainder of the lost half hour I assume is due to thread synchronization. Is that a reasonable assumption?

Ed F
7) Message boards : News : New N-Body (Message 57115)
Posted 1 Feb 2013 by EdwardPF
Post:
well at 2:00:00 the WU is 2.606% done ... I think that implies about 76.75Hours total time.

I ASSUME ;-) it will bomb out before then ( disk space exceeded) ... but we'll see ...

Ed F
8) Message boards : News : New N-Body (Message 57114)
Posted 31 Jan 2013 by EdwardPF
Post:
o.k. I continued to get

the procedure entry point pthread_attr_destroy could not be
located in the dynamic link library pthreadGC2_64.dll


so I reset, removed, and reconnected to the project.

I set pref's to n-body only and

" ... use at most 12.50% of the processors"

The n-body appears to be running fine with a est time of 56Hrs.

The 1 WU is consuming 46% of the system CPU resources.

Ed F

win-7 (64 bit core-7 "8" cpu's)
9) Message boards : News : New N-Body (Message 57102)
Posted 31 Jan 2013 by EdwardPF
Post:
All of my n-body WU are erroring out with a DLL error (can't find entry point - I believe)

Ed F
10) Message boards : Number crunching : Linux GPU: a huge memory leak?? (Message 57014)
Posted 24 Jan 2013 by EdwardPF
Post:
Note sure if this answers the q at all ... but ..
In windows-7 version the separation WUs run at 3,060 - 3,068 KB I have not seen a mem prob. yet.

Hopefully someone with the same OS can reply.

Ed F
11) Message boards : Number crunching : Very long run times on CPU Wu's (Message 56995)
Posted 23 Jan 2013 by EdwardPF
Post:
My 2 cents ...

Ya ... the nbody WU's are crazy long EST time (as high as 87,600:00:00 with actual run times at about 15 days (with 4 cpu's running it)

the "normal WE's" tend to give EST time at around 200Hrs but run in about 3.5 hrs ... the ps_separation_16 ... WU's seem to have more reasonable EST times.

Ed F
12) Message boards : Number crunching : N-Body blues (Message 56970)
Posted 21 Jan 2013 by EdwardPF
Post:
Hi!

I'm new here also ...

The nbody WU's are ranging up to 86000+ hours of est time. If you look they cane take up to 15-20 days actual ... if they don't error out (like mine did after 120Hrsw CPU ... and no credit:-)

The "regular W@H wu's seem to give an inflated est time (mine in the 200's of hrs) but seem to run in 4-6 hrs with a typical credit of about 215 cobbs.

I'm here for the "fun of it" ... actually ... I needed a heavy doubleprecision work load I could but on half of my CPU's ... this seems to work fine.


by the way ... the nbody WU's are multi-threaded and each WU will run on up to 4 CPU's at the same time


I'm not doing nbody WU's for now .... they seen unstable to me.

Ed F
13) Message boards : News : New N-Body Runs (Message 56958)
Posted 20 Jan 2013 by EdwardPF
Post:
My last 2 "big ones" errored out with disk space exceeded (boinc has it set to 50% - of 80Gb) and gave no credit (:-) ) the last one reported 0.00 sec CPU and 0.00 sec elapsed while in truth it was 42 Hrs elapsed (about 15%) and 120+ hrs CPU. so for now I'm not running nbody 'till I hear better reports.

Ed F
14) Message boards : Number crunching : What do you recommend (Message 56954)
Posted 19 Jan 2013 by EdwardPF
Post:
Well, it errored out after 40+ hours giving a reported time of 0.00 and now affigned to a new person (sorry) see WU 291619119.

It errored with disk exceeded ... whatever that means ... i expect it did not exceed the 40Gb boinc allows ... who knows.

so now I'b running 1 "normal" w@h ' till I UP boinc to 50%

Ed f
15) Message boards : Number crunching : What do you recommend (Message 56951)
Posted 18 Jan 2013 by EdwardPF
Post:
I have a screen shot from task manager showing 4 seti@home WU's running on CPU's 0,2,4,and 6 and 1 milkeyway@home WU (the nbody one) running on CPU's 1,3,5,and 7.

https://dl.dropbox.com/u/42596478/BOINC%20with%20MW%40H%20and%20S%40H%20running.jpg

I have separate copies of boinc running out of different dir's for s@h and m@h.

As you can see (if i did the URL correctly) the 4 s@h WU's are each consuming 100% of 4 CPU's while the MT nbody WU is consuming MOST of the 4 CPU's.


Fun huh!!

Ed F
16) Message boards : Number crunching : What do you recommend (Message 56950)
Posted 18 Jan 2013 by EdwardPF
Post:
PLEASE let us know if this works for you!!! My understanding is that it won't, but it could I guess. My understanding, I am just a cruncher like you, is that when you limit the cpu's, by the 12.5% in your case, then that is ALL Boinc will use. Using more goes outside the settings and that is not permitted, is my understanding. How I understand it works is that you are crunching along using all of your cpu's and in comes a unit that can use more than one core, so it uses as many cores as it wants, after all 100% are available and either the units on the other cores get paused or they don't start if they are not yet running. Now it could work like you are saying, I don't know for sure, so PLEASE come back and tell us how it is REALLY doing!


Yes it works just fine ( with BOINC 6.12.34). since my last post the WU has 23:24:00 elapsed time -6.74% finished - and 69:20:00 cpu time (a ratio of about 3:1 - remember the first 6 hours were run with 3 non-nbody WU's competing with it), Boinc has NOT scheduled any more WU's 'cause it's set to 12.5% BUT the 1 WU allowed by bounc (the nbody filling the 12.5% slot) is merrily running at at 46% usage according to task manager.

It seems to me that MT works just fine ( except it only uses 4 cpu's and not 8) and that bounc 6.12.34 does not know how schedule them.

When this one finishes (in 14 days) or errors out I have de-selected nbody 'till more info is available about it and boinc.

Ed F
17) Message boards : Number crunching : What do you recommend (Message 56939)
Posted 17 Jan 2013 by EdwardPF
Post:
I have a nbody running with a 6 Hour completion of 1% of total. At this rate it will take 600 Hrs to finish ... the deadline is 1/29/2013.

I could set "get no new tasks" and in about 5 min's the other 3 WU's will finish and the nbody will have 4 CPU's to run in.

BUT

If it errors out then I wont be doing any work.

Just how "stable" is nbody (today) ... should I let it use its full 4 CPU's or just let it run as-is and expect it to fail .... whenever?

Ed F

<edit> I set BOINC to 12.5% and reset to allow new tasks. This should allow 1 WU (1 CPU ) to run ... since the one running now uses 4 cpu's this should "fill" my -normal- use 50% of cpu's boinc quote. Once it finishes (or errors out) Boinc will allow 1 more WU "in". if it is nbody ... it'll run on 4 cpu's ... If its NOT nbody it will share 4 cpu's (using ~25% each) ... that's my guess.

Ed F
18) Message boards : Number crunching : A simple OS process accounting error?? (Message 56907)
Posted 16 Jan 2013 by EdwardPF
Post:
The two settings are (currently) set at 50% (schedule 4 cores) and 100% (all the time).

The first one should be telling BOINC how many WU's to allow to run - In my case that means that 4 WU's were running ... one of them MT so I Should have seen 100% CPU usage IFF the MT WU could use 5 or more cores. I only saw 88% busy ... I.E three ST WU's and one 4 thread WU (the nbody).

I set boinc to accept no new WU's and after the three ST WU's finished and the one MT WU was running I set the "use at most" to 100% (which shouldn't have affected anything) ... I made sure the "nbody" WU had affinity set to "all" ... and sure enough ... the computer was still 50% busy.

My conclusion ... "nbody" uses 4 sub threads and no more.

(Of course I may be wrong)

Ed F
19) Message boards : Number crunching : BOINC est time limit is 86600:00:00 (Message 56906)
Posted 16 Jan 2013 by EdwardPF
Post:
Mine errored out with disk limit exceeded 87.80 Mb exceeds 50.00 Mb.

The est time NEVER went down ... it was stuck at 87600:00:00

I don't see a 50.00 MB limit on disk space so I assume it's imposed by the WU.

Ed F
20) Message boards : Number crunching : A simple OS process accounting error?? (Message 56899)
Posted 15 Jan 2013 by EdwardPF
Post:
O.K. I still don't know ... I have 5 CPU;'s free and nbody is only using 4 of them and 1 is idle ...

Ed F


Next 20

©2019 Astroinformatics Group