Users April 2010

users@lists.einsteintoolkit.org

15 participants
22 discussions

GetComponents description
by Bruno Coutinho Mundim 20 Apr '10

20 Apr '10

Hi, reading the GetComponents description again, I noticed that the !NAME directive is not described there (at the end of the file all required and optional directives are described). Bruno.

1 0

test suite data files commits to svn
by Bruno C. Mundim 20 Apr '10

20 Apr '10

Hi, I have seen a few Thorns with a lot of test data files committed to the svn repository. I was wondering if that is really necessary. Committing only the parameter file that generates all those data files seems more natural to me and avoids unnecessary network traffic whenever we update the thorns. Besides it is cleaner and easier to spot what was really changed/updated. What do you think? Should we adopt as best practice not committing the data files themselves? and commit only the test suite parameter files? Thanks, Bruno.

5 7

Helping with moving thorns
by Erik Schnetter 19 Apr '10

19 Apr '10

As you probably saw on the Cactus Users' mailing list, we are moving certain thorns between repositories today, and we are switching the Cactus repositories from cvs to svn. While the thorns themselves (with a few exceptions) should not change at all, this change will require testing, and you can help us: - Wait for Frank's email that things are ready for testing - Frank will describe the new repository structure - Test checking out and building from new thorn locations - both anonymous and read/write access - test this for many architectures (both workstations and the big LONI/TeraGrid systems) - Look over CactusCode and EinsteinToolkit web sites - update download instructions and all thorn lists - use GetComponents instead of GetCactus - look at Utilities as well - Run test suites, compare output before and after - Begin running a few production simulations as well If you are helping, please keep in touch via frequent emails to the public mailing lists. -erik -- Erik Schnetter <schnetter(a)cct.lsu.edu> http://www.cct.lsu.edu/~eschnett/

5 8

Re: [Users] Fwd: test suite data files commits to svn
by Roland Haas 19 Apr '10

19 Apr '10

Hello all, >>> I have seen a few Thorns with a lot of test data files committed >>> to the svn repository. I was wondering if that is really necessary. >>> Committing only the parameter file that generates all those data >>> files >>> seems more natural to me and avoids unnecessary network traffic >>> whenever >>> we update the thorns. Besides it is cleaner and easier to spot what >>> was >>> really changed/updated. What do you think? Should we adopt as best >>> practice not committing the data files themselves? and commit only >>> the >>> test suite parameter files? If the concern is only to keep the commit emails small, then one could set the mime-type property of the files to binary before they are committed (see http://stackoverflow.com/questions/73797/how-do-i-tell-subversion-to-treat-…). svn propset svn:mime-type application/octet-stream filename.extension Suversion then no longer includes diffs in the commit emails for this file. This helps quite a bit when committing testsuite output. Yours, Roland

2 1

Re: [Users] [Developers] Test suit results
by Frank Loeffler 14 Apr '10

14 Apr '10

On Mon, Apr 12, 2010 at 07:59:37PM -0400, Roland Haas wrote: > Test suites which I can make run: Great! Thanks! > Test suites which I could not make run (successfully): > - ADMConstraints_Mask in ADMConstraints > - ADMConstraints_Physical in ADMConstraints > - ADMConstraints_Shift in ADMConstraints > - test_ADM in ADMConstraints They are all in ADMConstraints which IMHO has a very limited lifetime itself anyway, for production at least. We could probably run them including ADM, make sure they pass and include another testsuite using McLachlan. I would use this procedure for all teststuites which only fail because of ADM missing. > - checkpoint2 in AHFinderDirect > - needs LegoExcision LegoExcision will be part of the ET. I am not sure if it is generally available from the AEI, the relevant URL is https://svn.aei.mpg.de/numrel/Excision/LegoExcision/trunk/ > - Kerr-Cartoon in AHFinderDirect > - needs Cartoon2D Cartoon will also be part of the ET. It currently still in the Cactus BetaThorns. > - de_Sitter in Exact > - runs with ML+MoL but produces completely different results in > - eg. the lapse (which is set by Exact) > (AEIThorns/Exact/test/de_Sitter.par) > Missing: ADM_BSSN As far as I know ADM_BSSN is in AEI_Archive. It is probably time to archive this testsuite as well and replace it with the corresponding McLachlan version. It would be interesting to know if the testsuites which (only) need BSSN_MoL pass if that thorn is included. Can you please test that as well? > - bam in TwoPunctures > - needs bam_elliptic > (AEIThorns/TwoPunctures/test/bam.par) > Missing: BAM_ELLIPTIC I wouldn't just remove this and there is no real way to make it work publicly, so we may just have to accept that this doesn't run within the ET. > Missing: CARPETMASK This thorn is included in the ET thornlist. Maybe it got added after you build your configuration? Frank

3 3

GetComponents script and svn over https:
by Roland Haas 14 Apr '10

14 Apr '10

Hello all, for the testsuites I wanted to check out LegoExcision from https://svn.aei.mpg.de/numrel/Excision/LegoExcision/trunk/ (the http://svn.aei.mpg.de address did not work) and found that GetComponents seems to stall when trying to check the thorn out. Running with "-v" I get to see subversions prompt: ----------------------------------------------------------------- Checking out module: Excision/LegoExcision from repository: https://svn.aei.mpg.de/numrel/Excision/LegoExcision/trunk into: Testsuites/arrangements cwd = /localdata/rhaas3/Testsuites svn co https://svn.aei.mpg.de/numrel/Excision/LegoExcision/trunk arrangements/Excision/LegoExcision Error validating server certificate for 'https://svn.aei.mpg.de:443': - The certificate is not issued by a trusted authority. Use the fingerprint to validate the certificate manually! Certificate information: - Hostname: svn.aei.mpg.de - Valid: from Feb 23 16:02:12 2010 GMT until Feb 22 16:02:12 2015 GMT - Issuer: Max-Planck-Gesellschaft, DE - Fingerprint: 03:b4:e8:6e:d9:09:e9:93:72:e9:ff:fa:df:e4:2c:6d:1d:2a:e4:66 (R)eject or accept (t)emporarily? t Obviously one could fix this with the fixes that were proposed on the Cactus users list but it would be nice if GetComponets which might be the first thing new Cactus users use would display these prompts by default. Yours, Roland -- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from http://keys.gnupg.net.

2 1

Re: [Users] [Developers] CactusEinstein/ADM
by Allen Gabrielle 13 Apr '10

13 Apr '10

I don't remember but I don't think we had a long discussion on this. I suggest we leave it out. How about adding an EinsteinArchive repository to SVN to add move old thorns like this too, if a thorn is moved should add to the README when it was put there and why (e.g. Erik's text below). On Apr 12, 2010, at 2:02 PM, Erik Schnetter wrote: > I don't remember -- did we say we wanted to include thorn ADM, or > leave it out? > > It uses old APIs that we want to disable at some point (doesn't use > MoL, uses the old CalcTmunu interface, doesn't support higher order > differencing, is not parallelised with OpenMP, etc.). It also takes a > long time to compile, and I doubt that people will be interested in it. > > Should we leave it out of our Einstein Toolkit thorn list? > > If people are interested in an ADM code, then we can use our Kranc- > generated ADM code that is part of McLachlan (although I still don't > think people will be interested). > > -erik > > -- > Erik Schnetter <schnetter(a)cct.lsu.edu> http://www.cct.lsu.edu/~eschnett/ > > > > _______________________________________________ > Developers mailing list > Developers(a)einsteintoolkit.org > http://lists.einsteintoolkit.org/mailman/listinfo/developers

1 0

Test suit results
by Allen Gabrielle 13 Apr '10

13 Apr '10

Just resending Roland's mail to the users mail list (we removed the developers mail list for einsteintoolkit and decided to just have the users mail list for all mails). On Apr 12, 2010, at 6:59 PM, Roland Haas wrote: > Hello all, > > so I spend today working through the test suites that failed because > of missing thorns and tried to either make them work or find out why > they fail. Here are my results: > > ************************************************** > Test suites which I can make run: > ************************************************** > > + test_recover in IOHDF5 > + works once the thorns from CactusWave have been compiled in > (Recover from a WaveToy ID checkpoint file in HDF5 format) > Missing: WAVETOYC IDSCALARWAVEC > > + checkpoint in AHFinderDirect > + use BSSN_O2 since spatial_order=2 > (AEIThorns/AHFinderDirect/test/checkpoint.par) > Missing: BSSN_MOL > > + recover in AHFinderDirect > + use BSSN_O2 since spatial_order=2 > (AEIThorns/AHFinderDirect/test/recover.par) > Missing: BSSN_MOL > > + bowl-init in Exact > + removed ADM > (AEIThorns/Exact/test/bowl-init.par) > Missing: ADM > > + whisky_test_tov_ppm in Whisky > + use BSSN_O2 since second order is BSSN_MoL default > (TOV from TOVSolverC, ICN, Marquina, PPM, Polytrope) > Missing: BSSN_MOL > > + whisky_test_tov_ppm_eosg in Whisky > + use BSSN_O2 since second order is BSSN_MoL default > (TOV from TOVSolverC, ICN, Marquina, PPM, Polytrope, EOSG) > Missing: BSSN_MOL > > + whisky_test_tov_ppm_no_trp in Whisky > + use BSSN_O2 since second order is BSSN_MoL default > (TOV from TOVSolverC, ICN, Marquina, PPM, Polytrope) > Missing: BSSN_MOL > > + test_one_boost_max in Whisky_TOVSolverC > + remove evolution since itlast=0 > (CIGR/Whisky_TOVSolverC/test/test_one_boost_max.par) > Missing: BSSN_MOL > > + test_one_static_max in Whisky_TOVSolverC > + remove evolution since itlast=0 > (CIGR/Whisky_TOVSolverC/test/test_one_static_max.par) > Missing: BSSN_MOL > > + test_tov_carpet in Whisky_TOVSolverC > + remove evolution since itlast=0 > (Carpet unigrid, one TOV) > Missing: BSSN_MOL > > + test_two_av in Whisky_TOVSolverC > + remove evolution since itlast=0 > (CIGR/Whisky_TOVSolverC/test/test_two_av.par) > Missing: BSSN_MOL > > + test_two_max in Whisky_TOVSolverC > + remove evolution since itlast=0 > (CIGR/Whisky_TOVSolverC/test/test_two_max.par) > Missing: BSSN_MOL > > + CarpetWaveToyNewRecover_test_1proc in CarpetIOHDF5 > + works once the two thorns from CactusWave are compiled in > (Carpet/CarpetIOHDF5/test/CarpetWaveToyNewRecover_test_1proc.par) > Missing: IDSCALARWAVEC WAVETOYC > > + CarpetWaveToyRecover_test_1proc in CarpetIOHDF5 > + works once the two thorns from CactusWave are compiled in > (Carpet/CarpetIOHDF5/test/CarpetWaveToyRecover_test_1proc.par) > Missing: IDSCALARWAVEC WAVETOYC > > + CarpetWaveToyRecover_test_4proc in CarpetIOHDF5 > + works once the two thorns from CactusWave are compiled in > (Carpet/CarpetIOHDF5/test/CarpetWaveToyRecover_test_4proc.par) > Missing: IDSCALARWAVEC WAVETOYC > > + input_initial_data in CarpetIOHDF5 > + contains type in paramter file: $parfilename->$parfile works once > this > + is fixed and wavetoymol from carpetextra has been compiled in > (Carpet/CarpetIOHDF5/test/input_initial_data.par) > Missing: WAVETOYMOL > > ************************************************** > Test suites which I could not make run (successfully): > ************************************************** > > - ADMConstraints_Mask in ADMConstraints > - needs "static conformal" metric type > (Constraints for Schwarzschild with mask activated) > Missing: ADM > > - ADMConstraints_Physical in ADMConstraints > - should doable when replacing ADM->ML and adding MoL > - fails with nans in the constraints when I tried to use ML+MoL > however > (Constraints for Schwarzschild with physical metric) > Missing: ADM > > - ADMConstraints_Shift in ADMConstraints > - needs "static conformal" metric type > (Constraints for Schwarzschild with mask and shift activated) > Missing: ADM > > - test_ADM in ADMConstraints > - needs "static conformal" metric type > (Geodesically sliced black hole, ADM, Staggered leapfrog) > Missing: ADM > > - test_extract in Extract > - needs "static conformal" metric type > (Extract waveforms from BH) > Missing: ADM > > - kerr in IDAnalyticBH > - needs "static conformal" metric type > (1+log sliced Kerr black hole, ADM, double leapfrog) > Missing: ADM > > - test_bl in IDAnalyticBH > - needs "static conformal" metric type > (1+log sliced 4 BL black holes, ADM, double leapfrog) > Missing: ADM > > - test_misner in IDAnalyticBH > - needs "static conformal" metric type > (1+log sliced Misner black holes, ADM, double leapfrog) > Missing: ADM > > - test_axibrill in IDAxiBrillBH > - needs "static conformal" metric type > (Initial data for axisymmetric distored black hole) > Missing: ADM > > - test_axibrill_nostagger in IDAxiBrillBH > - needs "static conformal" metric type > (Initial data for axisymmetric distored black hole) > Missing: ADM > > - checkpoint2 in AHFinderDirect > - needs LegoExcision > (AEIThorns/AHFinderDirect/test/checkpoint2.par) > Missing: BSSN_MOL LEGOEXCISION > > - Kerr-Cartoon in AHFinderDirect > - needs Cartoon2D > (AEIThorns/AHFinderDirect/test/Kerr-Cartoon.par) > Missing: CARTOON2D > > - recover2 in AHFinderDirect > - needs checkpoint2 > (AEIThorns/AHFinderDirect/test/recover2.par) > Missing: BSSN_MOL LEGOEXCISION > > - test_ah in Dissipation > - needs "static conformal" metric type > (distorted BH, no rotation, fixed mesh refinement evolution > with carpet) > Missing: BSSN_MOL > > - test_ob in Dissipation > - needs "static conformal" metric type > (distorted BH, no rotation, fixed mesh refinement evolution > with carpet) > Missing: BSSN_MOL > > - de_Sitter in Exact > - runs with ML+MoL but produces completely different results in > - eg. the lapse (which is set by Exact) > (AEIThorns/Exact/test/de_Sitter.par) > Missing: ADM_BSSN > > - KS-tilted in Exact > - needs LegoExcision > (AEIThorns/Exact/test/KS-tilted.par) > Missing: LEGOEXCISION BSSN_MOL > > - Kerr-rotating-180-staggered in RotatingSymmetry180 > - needs LegoExcision > (AEIThorns/RotatingSymmetry180/test/Kerr-rotating-180- > staggered.par) > Missing: LEGOEXCISION BSSN_MOL > > - Kerr-rotating-180 in RotatingSymmetry180 > - needs LegoExcision > (AEIThorns/RotatingSymmetry180/test/Kerr-rotating-180.par) > Missing: LEGOEXCISION BSSN_MOL > > - Kerr-staggered in RotatingSymmetry180 > - needs LegoExcision > (AEIThorns/RotatingSymmetry180/test/Kerr-staggered.par) > Missing: LEGOEXCISION BSSN_MOL > > - Kerr in RotatingSymmetry180 > - needs LegoExcision > (AEIThorns/RotatingSymmetry180/test/Kerr.par) > Missing: LEGOEXCISION BSSN_MOL > > - Kerr-rotating-90-staggered in RotatingSymmetry90 > - needs LegoExcision > (AEIThorns/RotatingSymmetry90/test/Kerr-rotating-90- > staggered.par) > Missing: LEGOEXCISION BSSN_MOL > > - Kerr-rotating-90 in RotatingSymmetry90 > - needs LegoExcision > (AEIThorns/RotatingSymmetry90/test/Kerr-rotating-90.par) > Missing: LEGOEXCISION BSSN_MOL > > - bam in TwoPunctures > - needs bam_elliptic > (AEIThorns/TwoPunctures/test/bam.par) > Missing: BAM_ELLIPTIC > > - idpuncture in TwoPunctures > - needs PETSC which I don't know how to get to work > (AEIThorns/TwoPunctures/test/idpuncture.par) > Missing: IDPUNCTURE TATELLIPTIC TATJACOBI TATMG TATPETSC > > - qlm-bl in QuasiLocalMeasures > - fails with: Assertion `prolongation_order_time == 2' failed in > - Carpet/Initialize.cc when CarpetMask is added > - once Carpet::prolongation_order_time=2 is added to the parameter > file, > - then 33 files still differ significantly > (LSUThorns/QuasiLocalMeasures/test/qlm-bl.par) > Missing: CARPETMASK > > - outer-buffers in Carpet > - needs InterpToArray > (Carpet/Carpet/test/outer-buffers.par) > Missing: WAVEMOL IDWAVEMOL INTERPTOARRAY > > - waveinterp-1p in CarpetInterp > - needs InterpToArray > (Carpet/CarpetInterp/test/waveinterp-1p.par) > Missing: WAVEMOL IDWAVEMOL INTERPTOARRAY > > - waveinterp-2p in CarpetInterp > - needs InterpToArray > (Carpet/CarpetInterp/test/waveinterp-2p.par) > Missing: WAVEMOL IDWAVEMOL INTERPTOARRAY > > - test_BSSN_MoL_Carpet in CarpetRegrid > - uses "static conformal" metric_type > (1+log sliced black hole, ADM_BSSN, Iterative CN, Carpet with 2 > levels) > Missing: BSSN_MOL > > - test_BSSN_MoL_Carpet_keep in CarpetRegrid > - uses "static conformal" metric_type > (1+log sliced black hole, ADM_BSSN, Iterative CN, Carpet with 2 > levels) > Missing: BSSN_MOL > > - test_BSSN_MoL_Carpet in CarpetRegrid2 > - uses "static conformal" metric_type > (1+log sliced black hole, ADM_BSSN, Iterative CN, Carpet with 2 > levels) > Missing: BSSN_MOL > > ************************************************** > Test suites which I have to fully tested for one > reason or the other: > ************************************************** > > = extract_physical in Extract > = should doable when replacing ADM->ML and adding MoL > (Extract waveforms from BH) > Missing: ADM > > = bowl-evolution in Exact > = almost identical to bowl-init. In particular itlast=0! > (AEIThorns/Exact/test/bowl-evolution.par) > Missing: ADM > > = qlm-ks-boosted in QuasiLocalMeasures > = not tested (since qlm-bl fails) > (LSUThorns/QuasiLocalMeasures/test/qlm-ks-boosted.par) > Missing: CARPETMASK > > = qlm-ks-shifted in QuasiLocalMeasures > = not tested (since qlm-bl fails) > (LSUThorns/QuasiLocalMeasures/test/qlm-ks-shifted.par) > Missing: CARPETMASK > > = qlm-ks-tilted in QuasiLocalMeasures > = not tested (since qlm-bl fails) > (LSUThorns/QuasiLocalMeasures/test/qlm-ks-tilted.par) > Missing: CARPETMASK > > = qlm-ks in QuasiLocalMeasures > = not tested (since qlm-bl fails) > (LSUThorns/QuasiLocalMeasures/test/qlm-ks.par) > Missing: CARPETMASK > > = qlm-minkowski in QuasiLocalMeasures > = not tested (since qlm-bl fails) > (LSUThorns/QuasiLocalMeasures/test/qlm-minkowski.par) > Missing: CARPETMASK > > This reduces the number of unrunnable thorns from 55 to 32 plus the > ones that I cannot tell (7), ie 15 additional workable ones (and one > I must have forgotten to count somwhere). I don't think that I have > commit rights to all of the repositories to update/upload the > changed/new testsuites, so someone else will have to do so (the > parameter files are attached). Some more could be made to work by > using McLachlan in the ones using the "static conformal" > metric_type, though I have not been very successful with this. > > Yours, > Roland > > -- > My email is as private as my paper mail. I therefore support > encrypting > and signing email messages. Get my PGP key from http://keys.gnupg.net. > <test-parfiles.tar>_______________________________________________ > Developers mailing list > Developers(a)einsteintoolkit.org > http://lists.einsteintoolkit.org/mailman/listinfo/developers

1 0

about udb.pm/mdb.pm in simfactory
by Jian Tao 13 Apr '10

13 Apr '10

Hi Erik, I noticed that users are supposed to make the following changes to udb.pm. Edit udb.pm and * replace YOUR_LOGIN with your username on Queen Bee * YOUR(a)EMAIL.ADDRESS with your usual email address * YOUR_ALLOCATION with your project allocation on Queen Bee Although I believe that users who follow the instruction won't have trouble to locate where to change, I am wondering if it would be easier to have a file similar to Cactus parameter file to store pure key value pairs ? e.g. user="jtao" The same goes for mdb.pm. Regards, Jian

3 4

Download instructions use GetCactus
by Roland Haas 12 Apr '10

12 Apr '10

Hello all, the download instructions for Cactus Experts http://einsteintoolkit.org/info/documentation/experts.php refer to the GetCactus script but use a thornlist that only the new GetComponents script can parse. Yours, Roland -- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from http://keys.gnupg.net.

2 1

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Users April 2010