Code_saturne 1.4b trouble [install]

More
15 years 5 months ago #2377 by Claus
Replied by Claus on topic Re:Code_saturne 1.4b trouble [install]
I finished the calculation with two cores now B)

However; metis is a real pain for ecs to recognize - I have finally got it to recognize -lmetis (or somethink along the lines) but i CANNOT get it to find metis.h - I have uninstalled SCOTCH along with -devs. and pointed to .../opt/metis-4.0 where I even changed the structure to have an 'include' and 'lib' folder.

ECS compiles and runs fine nevertheless.

Closing input: "preprocessor_output"

Partitioning 16320 cells to 2 domains (METIS_PartGraphRecursive).

Total number of faces on parallel boundaries: 240
wall-clock time: 0.016352 s; CPU time: 0.010000 s

Number of cells per domain (histogramm):
[ 8160 ; 8160 ] = 2

Opening output: domain_number_2

Closing output: "domain_number_2"

Memory use and timing summary:

Total memory: 7.715 Mb

Wall-clock time: 0.023 s
CPU time: 0.010 s

.

.
| Partitioner finish |
`
'

Sort of related question: Do you know if Code_Aster (9.3) dependent on SCOCTH? My CA analysis is chucking away fine right now, even if I uninstalled SCOTCH.

Again, thank you for your invaluable help :)

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version

Please Log in or Create an account to join the conversation.

More
15 years 5 months ago #2378 by David Monfort
Replied by David Monfort on topic Re:Code_saturne 1.4b trouble [install]
Perfect! You're welcome :)

Quick note: it's normal for the Preprocessor not to use the metis.h header because, in its 4.0 version, it's doing quite a mess with the system headers :angry: The Preprocessor uses its own prototypes for the METIS functions it needs (this problem is not present in the 5.0pre2 version however).

As for Code_Aster, I have no idea except that I thought they we're using a sligthly modified version of METIS. You might ask on their forum.

Please Log in or Create an account to join the conversation.

More
15 years 3 months ago #2447 by Jake
Replied by Jake on topic Re:Code_saturne 1.4b trouble [install]
Where did you get the 1.4b version? All that's shown on the EDF site is 1.4.0 and 1.3.3, and both are next to impossible to compile on ubuntu 8.10 when trying to use syr_cs and syrthes.

Please Log in or Create an account to join the conversation.

More
15 years 3 months ago #2454 by Claus
Replied by Claus on topic Re:Code_saturne 1.4b trouble [install]
well, I had a look and it seems the files for 1.4b have changed to 1.4.0 - it stills reads 'development' version on the page, but maybe they are finishing it up as we speak :)

As for your problem with Syrthes, create a new thread and see if you can't get some help compiling it :)

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version

Please Log in or Create an account to join the conversation.

More
15 years 3 months ago #2482 by David Monfort
Replied by David Monfort on topic Re:Code_saturne 1.4b trouble [install]
Hello !

I've been to busy lately to post on the forum, but yes, we released to new version for the user club meeting (end of last year). I'll post in a different thread the announce.

Concerning the installation issue, what errors did you get?

Actually, I have to apologize because the 1.4.0 installer was completely useless... :blush: A new one should be uploaded soon, but I can provide it sooner if needed. Apart from that, no specific issue is known on Ubuntu systems (I have one myself).

syr_cs needs to be installed after SYRTHES. And the installation of SYRTHES is not provided by Code_Saturne installer, but is more or less straightforward. Please, create a specific thread if you encounter any related problem.

Regards

Please Log in or Create an account to join the conversation.

Moderators: catux
Time to create page: 0.144 seconds
Powered by Kunena Forum