Users July 2012

users@lists.einsteintoolkit.org

18 participants
30 discussions

Start a nNew thread

Einstein Toolkit Meeting Reminder
by Roland Haas 15 Jul '12

15 Jul '12

1 0

meeting minutes
by Roland Haas 09 Jul '12

09 Jul '12

Present were: Tanja, Roland, Ian, Yosef, Josh, Erik, Philipp * ET fall workshop location poll settled on RIT time is likely in weeks of Oct 22nd or 29th * required rebuilds - suspect it to be related to auto generated files - expect fix to be able to work along the lines of existing force-reconfig files - track down cause - create trac ticket * add paragraph to "How to contribute" [1] stating the patches are the preferred method of submitting changes, but that full files are acceptable as well. In both cases listing the base revision is a requirement. - Erik will follow up on ML patch * MHD paper status: - tests in flat space close to completion - needs some more text to describe figures - plotting scheme might need update - advection loop test stopped working recently. Will track down cause in GRHydro - Alv\'en test works now that ID is available - more complex test still show problems: -- unmagnetized Bondi with c2p fixes works for some time then fails with large piled up matter inside the horizon -- Josh will look into using excision - magnetized Bondi test is likely harder - TOV tests are done, look fine, plots need to be done - collapse runs are in progress - question as to which characteristic speed to use: -- Josh suggest to use the speed of light in moving frame (see emails on list) -- Bruno is working on deriving proper expressions -- re-running these runs takes several days * finish up MHD paper by time current grant finishes * make public announcement once MHD is well tested * MPI will be included as external thorn * Erik has made announcement on the users list * will require changes to thorns directly using MPI * will require realclean, reconfig, build cycle Yours, Roland [1] http://einsteintoolkit.org/documentation/contribute/

1 0

Einstein Toolkit Meeting Reminder
by Roland Haas 08 Jul '12

08 Jul '12

Hi Please consider joining the weekly Einstein Toolkit phone call at 10 am US central time on Mondays. As usual, you can find instructions how to join on the following web site: http://einsteintoolkit.org/community/support/ In short: the number is (+1) 225-578-4942 or (+1) 866-573-0359 and the conference id is 118682#. I will host this week's call, so hopefully everything will go smoothly. (Some) topics to discuss: - Einstein Toolkit Workshop Please choose your preferred location at http://www.doodle.com/htit7iktgmw3fuwy Right now there is a 3;1 preference for RIT - getting Cactus to rebuild from scratch when cctk.h (and related files) change - patch policy: how to encourage patches rather than full tarballs (eg. the UIUC Carpet patches that I believe no one dared look at so far and https://trac.einsteintoolkit.org/ticket/963) - MHD paper update? Please feel free to add to this list (which is now slanted towards what I am currently interested in). Roland Haas

1 0

ET(K) VirtualBox image and shared folders, revisited
by Scott Hawley 07 Jul '12

07 Jul '12

Hi, I'm revisiting the VirtualBox disk image from the workshop in April. Previously I had it set up to auto-mount my user directory from my Mac (/Users/shawley) into the VM's /home/etk-user/shawley. And it used to work (i.e. on the VM, you could see the ~shawley directory from my Mac). It's not doing that anymore. I'm not sure what changed. I had, in /et/rc.local on the VM, the line... mount -t vboxsf /Users/shawley /home/etk-user/shawley I've tried this ... with and without "sudo" first, and with and without full path specification. The "Auto Mount" flag was checked in "Settings" for the VirtualBox VM's "Settings" window. Some web forums threads suggest turning that off... Either way, I can no longer get my directory to be shared. Dennis/Frank/others? Can someone suggest a way to (re-)achieve this? Thanks, Scott

2 4

brill
by Comer Duncan 06 Jul '12

06 Jul '12

Hi, I finally got Cactus to build on my macbook pro today and ran static_tov with apparent success. Thanks for all those who helped. I have another question, admittedly naive. Namely, I want to run a Brill wave using McLaclan. Can you guys point me to the appropriate config to do this old test? I see that Gundlach (and maybe others) has made IDBrillData in EinsteinInitialData. So if I want to make a config which will run this problem presumably using ML_* and IDBrillData I need to create such, unless there is one around somewhere that I can use. Can someone(s) help me out? Thanks. Comer

2 1

Problem with a regridding.
by Petr Tsatsin 04 Jul '12

04 Jul '12

Hello, I've experienced a bug in new release ET_2012_05. After first regridding NaNs appear in gravitational fields. I have GRhydro at revision 374. Does anyone experienced the same problems? Thank you. -- Petr Tsatsin Graduate Student - Department of Physics Charles E. Schmidt College of Science Florida Atlantic University (561)-297-3386 cell: 561-692-2040 ptsatin(a)fau.edu, ptsatsin(a)gmail.com

2 1

Kraken compilation failure
by Petr Tsatsin 04 Jul '12

04 Jul '12

Hello, I still can't compile ET on kraken. Does anyone has the same problems? I got this error. /lustre/scratch/yakunin/ET_2011_10/Cactus/arrangements/EinsteinAnalysis/Multipole/src/utils.cc(22): catastrophic error: cannot open source file "hdf5.h" Thank you. -- Petr Tsatsin Graduate Student - Department of Physics Charles E. Schmidt College of Science Florida Atlantic University (561)-297-3386 cell: 561-692-2040 ptsatin(a)fau.edu, ptsatsin(a)gmail.com

2 1

question about checkout slowness
by Comer Duncan 03 Jul '12

03 Jul '12

Hi, I have cleaned out my old Cactus installation and am starting anew. I have run GetComponents for the development version. I began with no Cactus directory. Here is what I see: homelap-3:~ comerduncan$ ./GetComponents --noshallow -a https://svn.einsteintoolkit.org/manifest/trunk/einsteintoolkit.th ----------------------------------------------------------------- Checking out module: manifest from repository: http://svn.einsteintoolkit.org/manifest/trunk into: Cactus ----------------------------------------------------------------- Checking out module: Cactus from repository: http://svn.cactuscode.org/flesh/trunk into: Cactus as: . ----------------------------------------------------------------- Checking out module: simfactory2 from repository: https://svn.cct.lsu.edu/repos/numrel/simfactory2/trunk into: Cactus as: simfactory Now the above script was started at 14:17 on Friday 2 July. It is now 14:45 and the installation has not proceeded from the last line listed above. I have not nuked the attempt but am wondering if this is indicative of a problem somewhere at CCT? I surely am fine with standing in line if necessary. This just seems excessively long. I have looked at the Cactus directory and ls returns just: homelap-3:~ comerduncan$ ls ./Cactus total 144 -rw-r--r-- 1 comerduncan staff 968 Jul 3 14:17 CONTRIBUTORS -rw-r--r-- 1 comerduncan staff 19352 Jul 3 14:17 COPYRIGHT -rw-r--r-- 1 comerduncan staff 46292 Jul 3 14:17 Makefile drwxr-xr-x 16 comerduncan staff 544 Jul 3 14:17 src/ drwxr-xr-x 6 comerduncan staff 204 Jul 3 14:17 lib/ drwxr-xr-x 4 comerduncan staff 136 Jul 3 14:17 arrangements/ drwxr-xr-x 14 comerduncan staff 476 Jul 3 14:17 doc/ drwxr-xr-x 8 comerduncan staff 272 Jul 3 14:17 manifest/ so simfactory has just not made it (yet?). Has anyone seen this behavior before? I am running on my macbook pro running lion. I spent the morning working on getting macports populated with the list of required ports in the config file osx_lion_macports_gcc.cfg. So, I think I am ready to give building cactus a try. I have already successfully built the development version on my debian machine running squeeze and have used simfactory in the build and have successfully run the test on the tutorial page after making a couple of adaptations from queenbee to my local linux machine. Armed with that success I decided to give the macbook another try. This is where I am at the moment. Thanks for any help will be appreciated. Best, Comer

3 2

meeting minutes
by Roland Haas 02 Jul '12

02 Jul '12

Present were: Philipp, Erik, Bruno, Roland, Yosef, Josh, Tanja, Steve, Dennis * developers workshop locations for fall workshop are still open. Please choose your preferred location at http://www.doodle.com/htit7iktgmw3fuwy * cactuscode subversion server Steve will look into this from LSU (ticket # is https://trac.einsteintoolkit.org/ticket/969) * Roland will modify the failing utility to use a C/C++ main program and a Fortran subroutine to avoid the build failures * current stauts of the improved scheduler is that we have read/write statements and have plan to add interior/everything/boundary options to specify the region which the routine reads/writes (independently) * Steve and hari Krishnan are looking into creting a GUI, filter-pipeline like, interface to creaet schedule.ccl. Roland suggests that this might be useful if it shows a view of the whole schedule generated by Cactus while the user can modify a thorns schedule.ccl. Yours, Roland

1 0

Einstein Toolkit Meeting Reminder
by Roland Haas 01 Jul '12

01 Jul '12

Hi Please consider joining the weekly Einstein Toolkit phone call at 10 am US central time on Mondays. As usual, you can find instructions how to join on the following web site: http://einsteintoolkit.org/community/support/ In short: the number is (+1) 225-578-4942 or (+1) 866-573-0359 and the conference id is 118682#. I will host this week's call, so hopefully everything will go smoothly. (Some) topics to discuss: - Einstein Toolkit Workshop Time and location are not decided yet, the doodle poll for the time http://doodle.com/ritukz79hhuvhben is still open, and so far LSU (Frank) and RIT (Joshua) volunteered to host it, but no other opinions have been raised. I would like to make progress on this if at all possible so that people can plan flights etc. - subversion server woes - Fortran interfaces in particular to EOS_Omni and GRHydro - build failure of Riemann1D utility from GRHydro on some machines (hopper, kraken) - scheduling improvements (wrt GLOBAL, LOCAL, LOOP-LOCAL etc.) progress Please feel free to add to this list (which is now slanted towards what I am currently interested in). Roland Haas

1 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Users July 2012