Users

users@lists.einsteintoolkit.org

3 participants
3984 discussions

Minutes
by Frank Loeffler 09 Mar '15

09 Mar '15

Hi, Present were today: Frank Jonah Matt Peter Steve The main discussion revolved around using 'docker' as a way to deploy the ET, either at tutorials, or possibly HPC systems (although that would require more work and cooperation). Docker has its roots in cloud-computing (and is supported by several big players there), and was mentioned in a workshop Erik went to last week. It is a "virtual environment" instead of a "virtual machine". The difference is that unlike in a virtual machine, where an entire host is emulated, only a new 'software environment' is created, while that still runs under the same "host"-OS, using the same kernel. In the end, the user runs stuff (almost) like any other executable. This tight coupling between docker and the host-kernel is the reason for one possible disadvantage: it runs natively only in linux, and Erik mentions that is is possible to painlessly have it run also on Macs, using a tiny VM to run the linux kernel. Windows is, as far as everyone knew, not directly supported. The way to run docker executables there would be to run them within a regular virtual machine. Also, docker requires a daemon to run that needs root - which might require convincing arguments at HPC centers. In any case, Erik currently tries to have the ET release run in one of these containers, and will report about his success later on. Frank

1 0

Einstein Toolkit Meeting Reminder
by Frank Loeffler 09 Mar '15

09 Mar '15

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/ NOTE: Most of Louisiana changes to summer time this weekend, while other regions might not. In these cases, "usual meeting time" means one hour earlier than usual. In short: the number is (+1) 225-578-4942 and the conference id is 118682#. Frank Löffler

3 2

Re: [Users] is interpolation of grid arrays supported?
by Roland Haas 08 Mar '15

08 Mar '15

Hello Yosef, > I should have clarified this to begin with, I was using a single > processor. Well in that case my comment clearly does not apply. > It seems that somehow the extent of the grid array is lost. > The grid, as seen by AEILocalInterp is only 9 elements > long. This may be unrelated: but are you using a development version of the flesh? I fixed a bug in the parameter file parser that affects things like: size=\ (NullGrid::N_ang_pts_inside_eq+2*(NullGrid::N_ang_ev_outside_eq+NullGrid::N_ang_stencil_size)),\ (NullGrid::N_ang_pts_inside_eq+2*(NullGrid::N_ang_ev_outside_eq+NullGrid::N_ang_stencil_size)) in interface.ccl (the fix is in 635f90e "Cactus: fix access to deallocated temporary variable in piraha"). This would affect the size of the array in memory as well though, so I would expect it to fail before. Otherwise, I am as baffled as you are. 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. -- 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.

1 0

Re: [Users] [Einstein Toolkit] #1712: Retire snprintf.c
by Roland Haad 06 Mar '15

06 Mar '15

Gcc warns. But I get so many warnings due to variable shadowing and fortran subroutine calls without interface that I only see them if I hunt for them. Its not as if I recompile the flesh very often. :-) We could make the warning an error and that should help but may break some code tgat does not eg declare printf. Yours Roland ----- Original Message ----- From: Einstein Toolkit <trac-noreply(a)einsteintoolkit.org> Sent: 2015/03/06 - 21:19 To: null Subject: Re: [Einstein Toolkit] #1712: Retire snprintf.c > #1712: Retire snprintf.c > -----------------------+---------------------------------------------------- > Reporter: eschnett | Owner: > Type: task | Status: review > Priority: minor | Milestone: > Component: Other | Version: development version > Resolution: | Keywords: > -----------------------+---------------------------------------------------- > > Comment (by eschnett): > > Regarding {{{strdup}}}: Ouch. Hello 1972, nice to meet you. > > Isn't there a gcc warning that requires prototypes for all functions? I > think we should be able to enable this safely. > > -- > Ticket URL: <https://trac.einsteintoolkit.org/ticket/1712#comment:6> > Einstein Toolkit <http://einsteintoolkit.org> > The Einstein Toolkit > -- 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.

1 0

meeting minutes for 2015-03-02
by Roland Haas 02 Mar '15

02 Mar '15

present: Matt, Peter, Frank, Roland, Ian, Zach, Josh, Yosef Rewrite branch: no news C++ code in GRHydro: * about a factor of 2 speedup compared to F90 code * not clear what the cause of the speedup is, possibly just the fact that all if statements were replaced by template parameters * Roland to send to Peter a parameter file for a supernova type run to check for hot spots in the code Yours, Roland

1 0

Einstein Toolkit Meeting Reminder
by Frank Loeffler 27 Feb '15

27 Feb '15

1 0

Meeting Minutes for 2015-02-23
by Roland Haas 23 Feb '15

23 Feb '15

Present: Roland, Ian, Peter, Steve, Yosef, Philipp, TwoPunctures: * fix is accepted. impact is unclear, should only affect convergence speed * Yosef suspects that this affects the direction that the solver takes since it only affects the preconditioner McLachlan rewrite branch: * latest news: Erik wanted to run benchmarks, Ian will also try but is very busy, same for Peter MHD constrained transport: * Philipp, and the Caltech group have started discussion of constrained transport algorithms but have not reached a decision what to do * current best practice is to add some resistivity to the computation of the electric field, this is consistent with constrained transport. The numerical coefficient is 1e-3 -- 1e-4 . Currently this is under heavy development and not even in Zelmani yet. TOVSolver tying to GRHydro: * Roland will go ahead and remove the linkage Hydro in Kranc: * Ian Hawke (Southampton) is interested in doing a full timestep evolution scheme (some predictor/corrector) in Kranc Yours, Roland

1 0

Einstein Toolkit Meeting Reminder
by Frank Loeffler 20 Feb '15

20 Feb '15

1 0

Configurations for PRACE machines
by Ian Hinder 17 Feb '15

17 Feb '15

Hi all, We are interested in running on some European PRACE machines that we have not used before. Does anyone have SimFactory (or other) Cactus configurations for Curie, Hornet or Mare Nostrum? We would like to get these working with the ET, and then commit their definitions to SimFactory. -- Ian Hinder http://members.aei.mpg.de/ianhin

1 0

No Einstein Toolkit Meeting
by Frank Loeffler 13 Feb '15

13 Feb '15

Hi, There will not be an Einstein Toolkit meeting coming Monday. The next meeting will be in about one week on Monday, February 23th, at the usual time. Frank Löffler

1 0

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Users