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/
I short: the number is (+1) 225-578-4942 or (+1) 866-573-0359 and the conference id is 118682#.
We should finalize the time for the fall workshop, for people to make their travel plans. Currently the week of Oct 22nd is proposed and nobody objected. However, exact dates aren't decided yet. That, and topic(s) to work on can be entered (by anyone) in the wiki:
https://docs.einsteintoolkit.org/et-docs/ET_Workshop_Fall_2012
The CIGR award will come to an end this year. This is the grand that funded the toolkit for the past three years. While we got funding for another three years, a final report will need to be written. I will contact anyone I would expect to contribute in person, but it would also be nice to get input from the toolkit community. Please spare 5 minutes and think about how the toolkit helped your research. Write a couple of sentences and send them to us. If you would like to go that far you could include a list of publications that benefited from the toolkit.
Also, in case you are using parts of the toolkit and are not mentioned on http://einsteintoolkit.org/about/members/ yet, please consider registering. It is not attached to any strings, it only helps the maintainers to better justify what they do.
As always please speak up with any issue, question or story you have connected to your work and the toolkit.
Activity last week: 30 issues changed and still open, 30 issues closed, e.g.: - GRHydro received major updates - EOS table can be readon single process and it then broadcast - added option checkpoint_every_divisor - Don't log in multiple times for sim sync - various fixes and improvements to checkpointing/restarts
Frank Loeffler
Hello all,
As always please speak up with any issue, question or story you have connected to your work and the toolkit.
I'd like to add the current issue with McLachlan/Kranc and vectorization to the topics for the phone call.
Yours, Roland
Hi,
could you tell a little more about checkpoint_every_divisor? What and where is it?
Cheers, Luca
On 8/13/12 10:14 AM, Frank Loeffler wrote: [...]
Activity last week: 30 issues changed and still open, 30 issues closed, e.g.:
- GRHydro received major updates
- EOS table can be readon single process and it then broadcast
- added option checkpoint_every_divisor
- Don't log in multiple times for sim sync
- various fixes and improvements to checkpointing/restarts
Hello Luca,
could you tell a little more about checkpoint_every_divisor? What and where is it?
It is an option of CarpetIOHDF5. Right now there is an option checkpoint_every that already takes a number iterations as its frequency. checkpoint_every works the way all other _every options work in the core IO thorns (ie CarpetIOASCII, CarpetIODF5, CactusIOASCII etc but not eg Multipole) and triggers checkpointing if the difference between cctk_iteration and the iteration of the last checkpoint is larger than the given frequency. checkpoint_every_divivisor writes a checkpoint whenever cctk_iteration % checkpoint_every_divider == 0 (in C notation in Fortran this would be the .mod. operator I think).
If you never change the checkpointing frequency and your termination checkpoints when you run out of walltime are also aligned with checkpoint_every (not sure why they should though), then there is no difference between the two. Otherwise checkpoint_every_divisor can be used to eg. get a checkpoint at a certain iteration number or to make sure that the periodic checkpoints are written on coarse steps (it does nothing for the termination checkpoints).
Yours, Roland
users@lists.einsteintoolkit.org
-
Frank Loeffler -
Luca Baiotti -
Roland Haas