Gravitational Wave search O2 All-Sky Tuning ("O2AS20-500T")

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4267
Credit: 244933143
RAC: 16332
Topic 213199

On the weekend we began to send out the first tasks of the first search for Continuous Gravitational Waves in "O2" data, i.e. data from the second Observation Run of the LIGO detectors. This "Tuning" run will be a short one (targeted for about a month) where we will test and tune the application and setup for the actual analysis run, which will follow immediately (or will even be started before the tuning run is fully finished).

As the name indicates, this will be an all-sky search, covering the frequencies 20-500Hz. For the tuning run, however, we will focus on a few smaller frequency bands in that range.

Currently this is marked as a "Beta test" application (meaning you have to accept "Beta work" in your preferences to get it), but this is meant to be changed soon (today or tomorrow). We still have trouble building the OSX version of the application, so there currently is no such for Macs, but that should get fixed within this week.

BM

Darren Peets
Darren Peets
Joined: 19 Nov 09
Posts: 37
Credit: 98002100
RAC: 30544

On my system, it looks like

On my system, it looks like they take ~40% longer than the most recent chunks of LAT data, which is about what I recall for O1-High (although I guess LAT can be highly variable).

Consider tuning the number of "T"s in "Gravittational"...

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4267
Credit: 244933143
RAC: 16332

Thanks!

Thanks!

BM

MarkJ
MarkJ
Joined: 28 Feb 08
Posts: 437
Credit: 137331180
RAC: 18170

Are we expecting similar

Are we expecting similar memory usage with the O2 run to what we had in the O1 run? (ie up to 1500MB per work unit).

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4267
Credit: 244933143
RAC: 16332

Actually there were three

Actually there were three "O1" runs ("O1AS" (all-sky), "O1MD" (multi-directed) and the recent "O1Spot" (single-directional)).The "directional" searches were looking up to much higher frequencies than the "all-sky" search, that's where most of the memory usage came from.

I think the O2 all-sky run will be in the same order of magnitude than the O1 all-sky run (250-300MB), possibly up to a factor of two higher (600MB). Our model of memory-usage per data is certainly not perfect, we'll use the "tuning" run to improve this as well.

BM

Thyme Lawn
Thyme Lawn
Joined: 13 Jun 15
Posts: 7
Credit: 12137755
RAC: 0

9 tasks successfully

9 tasks successfully completed and 1 failed with exit status 13 on my i7 Windows 10 Pro system so far.

Runtime is comparable to that reported by Darren Peets (~9 hours and 1000 credits for O2, ~5 hours and 693 credits for FGRP5).

The failure ran for 4 hours and was reissued to and successfully completed by a Linux system (WU h1_0117.90_O2C01Cl1In1__O2AS20-500T_118.00Hz_83.

End of stderr on my system:

298..............ERROR: Stepped outside the coarse grid (79:14971:15049:15461)!

2018-02-05 22:12:33.5469 (12396) [CRITICAL]: ERROR: MAIN() returned with error '13'
FPU status flags:  COND_2 PRECISION
2018-02-05 22:12:33.5469 (12396) [normal]: done. calling boinc_finish(13).
22:12:33 (12396): called boinc_finish

And on the Linux system:

685........c

...........
2018-02-06 12:21:05.2553 (15172) [normal]: Finished main analysis.
2018-02-06 12:21:05.2557 (15172) [normal]: Recalculating statistics for the final toplist...
2018-02-06 12:24:56.9140 (15172) [normal]: Finished recalculating toplist statistics.
FPU status flags:  COND_1 PRECISION
2018-02-06 12:24:57.2983 (15172) [normal]: done. calling boinc_finish(0).
12:24:57 (15172): called boinc_finish

The first task from the WU was sent to another Linux system and immediately failed with exit status 1 and the following in stderr (all 9 O2 tasks on that system have failed in the same way):

einstein_O2AS20-500T_0.01_x86_64-pc-linux-gnu: option `--getMaxFperSeg' was set more than once!

Usage: einstein_O2AS20-500T_0.01_x86_64-pc-linux-gnu [-h|--help] [-v|--version] [@<config-file>] [--log] [--semiCohToplist] --DataFiles1 [--skyRegion] [--numSkyPartitions] [--partitionIndex] [--skyGridFile] [--dAlpha] [--dDelta] [-f|--Freq] [--dFreq] [-b|--FreqBand] [--f1dot] [--df1dot] [--f1dotBand] [--f2dot] [--df2dot] [--f2dotBand] [--f3dot] [--df3dot] [--f3dotBand] [--peakThrF] [-m|--mismatch1] [--gridType1] [--metricType1] [-g|--gammaRefine] [-G|--gamma2Refine] [-o|--fnameout] [--fnameChkPoint] [-n|--nCand1] [--printCand1] [--refTime] [--ephemEarth] [--ephemSun] [--minStartTime1] [--maxStartTime1] [--printFstat1] [--assumeSqrtSX] [--nStacksMax] [-T|--tStack] [--segmentList] [--recalcToplistStats] [--loudestSegOutput] [--computeBSGL] [--Fstar0sc] [--oLGX] [--getMaxFperSeg] [--SortToplist] [--FstatMethod] [--FstatMethodRecalc]

2018-02-03 15:22:42.9481 (2622) [CRITICAL]: ERROR: MAIN() returned with error '1'
FPU status flags:
2018-02-03 15:22:42.9481 (2622) [normal]: done. calling boinc_finish(1).
15:22:42 (2622): called boinc_finish

 
 

"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4267
Credit: 244933143
RAC: 16332

The first tasks instantiated

The first tasks instantiated from the first <500 workunits had indeed been generated with a broken command-line, so they failed immediately, but without wasting any computing time. This was fixed; all tasks generated after Saturday noon (UTC) should not exhibit this problem.

"ERROR: Stepped outside the coarse grid" is pretty unusual, in particular if this doesn't happen on a different host. I'll take a look and try to reproduce it.

BM

DoctorNow
Joined: 22 Jan 05
Posts: 12
Credit: 27390545
RAC: 0

Hi! I don't see an app to

Hi!

I don't see an app to select in my account, Beta option is ACTIVATED but since there MUST be at least one app selected I only get tasks from this instead of the Beta ones.

Is there a way to get tasks for this app which I probably don't see?

Life is Science, and Science rules. To the universe and beyond
Proud member of BOINC@Heidelberg

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4267
Credit: 244933143
RAC: 16332

Thanks for the reminder.

Thanks for the reminder. Should be fixed (i.e. available in your preferences) now.

BM

mmonnin
mmonnin
Joined: 29 May 16
Posts: 291
Credit: 3229897254
RAC: 1093477

Bernd Machenschalk

Bernd Machenschalk wrote:
Thanks for the reminder. Should be fixed (i.e. available in your preferences) now.

 

Thank you, that helps.

Betreger
Betreger
Joined: 25 Feb 05
Posts: 987
Credit: 1421910223
RAC: 774075

Ah, that explains why haven't

Ah, that explains why haven't been able to get any.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.