Test "Binary neutron star" example

Issue #2176 open
Roland Haas created an issue

Before each release, check that http://einsteintoolkit.org/gallery/bns/index.html still works and produces correct output.

Then upload the results file to https://bitbucket.org/einsteintoolkit/www/downloads/

Keyword: None

Comments (70)

  1. Roland Haas reporter

    Still images can be created from movie frames produced by make_movie.sh script using:

    # create still images for website
    cp rho_000000000.png rho.png
    convert -extract 300x225+214+195 rho.png rho_4x3.png
    

    currently (as of hash 6db552c only rho_4x3.png is actually used).

  2. Beyhan Karakaş

    The gallery example “binary neutron star” was last successfully tested on 12 November 2020 and website was updated.

  3. Roland Haas reporter

    Uh right. The svn server svn.cct.lsu.edu was recently moved behind a firewall and is no longer publicly accessible. Though right now even when trying to access it from inside of LSU IP space (which worked fine yesterday) I get a error 500 internal server error.

    @Peter Diener @Steven R. Brandt is there a LSUThorns Bitbucket or github place where the thorns that wwere in https://svn.cct.lsu.edu/repos/numrel could be hosted? From the email that I received from IT support it seemed to me that the svn server would not be made public again (“svn access is restricted to internal users now.”).

  4. Roland Haas reporter

    For now I have put a copy of the required thorn (only the most recent revision which is all I had a copy of) to:

    https://bitbucket.org/rhaas80/lsuthorns/src/master/

    and you should be able to use it with a GetComponents line like so:

    # LSU Thorns
    !TARGET   = $ARR
    !TYPE     = git
    !URL      = https://bitbucket.org/rhaas80/lsuthorns.git
    !REPO_PATH= $2
    !CHECKOUT =
    LSUThorns/NSTracker
    

  5. Atul Kedia

    Yes, that seems to be working for me.

    Should I temporarily edit the thornlist on BNS website as well? Or we can discuss in the meeting tomorrow.

  6. Roland Haas reporter

    I would rather not point anything to my Bitbucket repo since that repo really just contains the single revision of the code. The svn server seems down right now so even from within LSU one cannot get a fresh checkout with all version history for git. Also since the thorn was developed at LSU I would rather not take ownership of it in my personal Bitbucket account.

  7. Atul Kedia

    Thanks for that.

    For the record, Presync enabled gives the following error:

    Required read for SPHERICALSURFACE::sf_area[0] (rl=0) not satisfied. Have Nowhere and require Everywhere missing Everywhere at the start of routine CarpetMask::CarpetSurfaceSetup. Current valid state:
    Valid entries: SPHERICALSURFACE::sf_area[0] (rl=0) Nowhere.

    Looks like an easy fix. I have no experience with both this thorn and presync to know how that can be done.

  8. Atul Kedia

    I reran the test on the Lorentz release version and am getting an error after certain time steps.

    Traceback (most recent call last):
    File "./simfactory/bin/../lib/sim.py", line 149, in <module>
    main()
    File "./simfactory/bin/../lib/sim.py", line 145, in main
    CommandDispatch()
    File "./simfactory/bin/../lib/sim.py", line 107, in CommandDispatch
    module.main()
    File "/afs/crc.nd.edu/user/a/akedia/etk_lorentz/Cactus/repos/simfactory2/lib/sim-manage.py", line 398, in main
    CommandDispatch()
    File "/afs/crc.nd.edu/user/a/akedia/etk_lorentz/Cactus/repos/simfactory2/lib/sim-manage.py", line 377, in CommandDispatch
    exec("command_%s()" % command)
    File "<string>", line 1, in <module>
    File "/afs/crc.nd.edu/user/a/akedia/etk_lorentz/Cactus/repos/simfactory2/lib/sim-manage.py", line 185, in command_create_run
    command_run()
    File "/afs/crc.nd.edu/user/a/akedia/etk_lorentz/Cactus/repos/simfactory2/lib/sim-manage.py", line 222, in command_run
    restart.userRun(simulationName)
    File "/afs/crc.nd.edu/user/a/akedia/etk_lorentz/Cactus/repos/simfactory2/lib/simrestart.py", line 1014, in userRun
    errfh.write(err_txt)
    IOError: [Errno 5] Input/output error

    The two times I ran the test on Lorentz gave the same error although at different timesteps (t = 460 and t = 920)

    I last tested the simulation on 17/May based on the master branch of ET and the run looked good then.

    I think this is an issue on my machine’s end and I have some ideas for fixing it, and will be trying them until I post here again.

  9. Atul Kedia

    When building ET using nsnstohmns.th I get the error

    Processing CCL files
    
    CST error in /home/asksma/etk_submit1/Cactus/repos/flesh/lib/sbin/interface_parser.pl (at 637)
      -> NSTracker (thorn NSTracker) inherits from HYDRO_ANALYSIS
        No thorn in your current ThornList implements HYDRO_ANALYSIS
        Either remove NSTracker, or add a thorn to your
        ThornList implementing HYDRO_ANALYSIS
        Available thorns in arrangements directory implementing HYDRO_ANALYSIS:
          EinsteinAnalysis/Hydro_Analysis
    

    and the same error for CactusNumerical/SphericalSurface

    Does anyone know why this is happening? Looks like it is trying to build NSTracker before building thorns from einsteintoolkit.th.

  10. Roland Haas reporter

    You need to pass nsnstohms.th to GetComponents first which should download the missing (non-ET) thorns.

  11. Atul Kedia

    I tried getcomponents of either 1. just nsnstohmns.th or 2. both nsnstohmns.th and einsteintoolkit.th. Following which I do the simfactory build nsnstohmns.th command to get the error message in both cases.

  12. Roland Haas reporter

    Hmm. Worksforme. These are the commands that I entered:

    curl -kLO https://raw.githubusercontent.com/gridaphobe/CRL/master/GetComponents
    chmod a+x GetComponents
    ./GetComponents --root ET_NSNSToHMNS --parallel --shallow http://einsteintoolkit.org/gallery/bns/nsnstohmns.th
    cd ET_NSNSToHMNS/
    ./simfactory/bin/sim setup-silent
    ./simfactory/bin/sim build --thornlist thornlists/nsnstohmns.th
    

    Were there any warnings eg from GetComponents about not being able to check out all thorns?

  13. Atul Kedia

    I am doing almost the same procedure. I get the message 304 components checked out successfully after checkout, and the ones mentioned in the error also came through without errors.

  14. Roland Haas reporter

    Let me be probe a bit more here. Which is he exact thorn list that you pass to simfacotyr (not GetComponents)? Could you maybe attach it to the ticket (otherwise we are mostly just guessing)? I am asking because you must not use the file nsnstohmns.th that GetComponents dropped into the directory it ran in (so simfactory/bin/sim build --thornllist ../nsnstohmns.th is incorrect) and you must instead use the one GetComponents writes to thornlists/nsnstohmns.th which has the !INCLUDE directive resolved. The one downloaded looks like so:

    # Component list for the Einstein Toolkit <http://einsteintoolkit.org/>
    # $Revision$
    # $Date$
    # $HeadURL$
    
    !CRL_VERSION = 1.0
    
    !INCLUDE = https://bitbucket.org/einsteintoolkit/manifest/raw/ET_2021_11/einsteintoolkit.th
    
    !TARGET   = $ARR
    !TYPE     = git
    !URL      = https://bitbucket.org/knarrff/nstracker.git
    !REPO_PATH = ../../arrangements/$1/../../repos/nstracker
    !CHECKOUT =
    LSUThorns/NSTracker
    

    and does (for me at least) produce an error message:

    ------------------------------------------------------
    There were 2 errors during execution of the CST
    These must be corrected before compilation can proceed
    ------------------------------------------------------
    
    
    
    ------------------------------------------------------
    Warnings were generated during execution of the CST
    ------------------------------------------------------
    
    
    CST error in /data/rhaas/postdoc/gr/cactus/ET_NSNSToHMNS/repos/flesh/lib/sbin/interface_parser.pl (at 637)
      -> NSTracker (thorn NSTracker) inherits from HYDRO_ANALYSIS
         No thorn in your current ThornList implements HYDRO_ANALYSIS
         Either remove NSTracker, or add a thorn to your
          ThornList implementing HYDRO_ANALYSIS
         Available thorns in arrangements directory implementing HYDRO_ANALYSIS:
            EinsteinAnalysis/Hydro_Analysis
    
    
    CST error in /data/rhaas/postdoc/gr/cactus/ET_NSNSToHMNS/repos/flesh/lib/sbin/interface_parser.pl (at 637)
      -> NSTracker (thorn NSTracker) inherits from SPHERICALSURFACE
         No thorn in your current ThornList implements SPHERICALSURFACE
         Either remove NSTracker, or add a thorn to your
          ThornList implementing SPHERICALSURFACE
         Available thorns in arrangements directory implementing SPHERICALSURFACE:
            CactusNumerical/SphericalSurface
    

    when passed to simfactory.

  15. Atul Kedia

    Thanks Roland, that was it. I had forgotten that the thorn in Cactus/thornlists needs to be passed instead of the ../nsnstohmns.th.

  16. Bing-Jyun Tsao

    For 2022_11_ET BNS gallery test: when building from the thornlist ./simfactory/bin/sim build bns --thornlist thornlists/nsnstohmns.th, I got the following error when compiling the ExternalLibraries/LORENE thorn.

    patching file Lorene/C++/Source/Bin_hor/Makefile
    patch: **** can't open file Lorene/C++/Source/Bin_hor/Makefile_lib : Too many open files
    make[3]: *** [/home/bt22824/2022_11_ETK/Cactus/arrangements/ExternalLibraries/LORENE/src/make.code.deps:14: /home/bt22824/2022_11_ETK/Cactus/configs/bns/scratch/done/LORENE] Error 2
    make[2]: *** [/home/bt22824/2022_11_ETK/Cactus/lib/make/make.thornlib:113: make.checked] Error 2
    make[1]: *** [/home/bt22824/2022_11_ETK/Cactus/lib/make/make.configuration:179: /home/bt22824/2022_11_ETK/Cactus/configs/bns/lib/libthorn_LORENE.a] Error 2
    

    For reference, here are the commands I used to download the current master version of ET

    curl -kLO https://raw.githubusercontent.com/gridaphobe/CRL/master/GetComponents
    chmod a+x GetComponents
    ./GetComponents --parallel --shallow https://bitbucket.org/einsteintoolkit/manifest/raw/master/einsteintoolkit.th
    ./GetComponents http://einsteintoolkit.org/gallery/bns/nsnstohmns.th
    

    This actually also occurs in an older ETK version I had (2021_05) when compiling the new thornlist nsnstohmns.th from 2022_05 release.

  17. Roland Haas reporter

    Are you compiling on a cluster (which one)? In that case, it could be a limit imposed by the admins. Would you mind trying with ./simfactory/bin/sim build -j1 bns to use only a single parallel make process? Also if you could provide the output of ulimit -a that would help.

  18. Bing-Jyun Tsao

    It seems like the issue only occurs when building on a local machine, but not on a cluster (i.e. Frontera, Expanse). On a local machine, the same error from Lorene still shows up even when restricting to a single make process.

    ./simfactory/bin/sim build -j1 bns --thornlist thornlists/nsnstohmns.th
    

    Since the issue only occurs on the local machine, perhaps some compiling packages on the machine should be updated to resolve the issue (linux machine last updated last year around October from https://github.com/nds-org/jupyter-et/blob/master/tutorial-server/notebooks/CactusTutorial.ipynb).

  19. Roland Haas reporter

    On the local machine (with the failure), what is the output of ulimit -a? “local” here means all local file systems, not anything (eg $HOME) mounted via NFS? (trying to understand what may be going on).

  20. Bing-Jyun Tsao
    core file size          (blocks, -c) 0
    data seg size           (kbytes, -d) unlimited
    scheduling priority             (-e) 0
    file size               (blocks, -f) unlimited
    pending signals                 (-i) 126665
    max locked memory       (kbytes, -l) 65536
    max memory size         (kbytes, -m) unlimited
    open files                      (-n) 1024
    pipe size            (512 bytes, -p) 8
    POSIX message queues     (bytes, -q) 819200
    real-time priority              (-r) 0
    stack size              (kbytes, -s) 8192
    cpu time               (seconds, -t) unlimited
    max user processes              (-u) 126665
    virtual memory          (kbytes, -v) unlimited
    file locks                      (-x) unlimited
    

    Here is the output with ulimit -a. You are correct, this is actually ran on a computer that is shared among our group members.

  21. Roland Haas reporter

    Hmm, not any different from my laptop (which also shows the 1024 open files restriction).

    real-time non-blocking time  (microseconds, -R) unlimited
    core file size              (blocks, -c) 0
    data seg size               (kbytes, -d) unlimited
    scheduling priority                 (-e) 0
    file size                   (blocks, -f) unlimited
    pending signals                     (-i) 78729
    max locked memory           (kbytes, -l) 2526916
    max memory size             (kbytes, -m) unlimited
    open files                          (-n) 1024
    pipe size                (512 bytes, -p) 8
    POSIX message queues         (bytes, -q) 819200
    real-time priority                  (-r) 0
    stack size                  (kbytes, -s) 8192
    cpu time                   (seconds, -t) unlimited
    max user processes                  (-u) 78729
    virtual memory              (kbytes, -v) unlimited
    file locks                          (-x) unlimited
    

  22. Bing-Jyun Tsao

    BNS gallery example (bns-20221027) is updated on the website.

    Notes on running BNS gallery example (10-27):
    - Gnuplot issue with epstopdf and convert: need to update gnuplot and allow read/write access with ImageMagick.
    - Add python script to plot Psi4 using kuibit instead (plot_quibit.py in script.tar.gz).
    - Compiling issue with LORENE: might not have the latest version of cmake, resolved by running on Expanse and Frontera.

  23. Log in to comment