OGS5 and PETSc

Luca Urpi

unread,

May 22, 2017, 10:25:54 AM5/22/17

to ogs-...@googlegroups.com

Dear OGS developers,

although OGS5 is at the end of its development (or am I wrong?) may you update the page

https://docs.opengeosys.org/docs/devguide5/getting-started/build-configuration#panel3-2

adding the information that to compile with PETSc not only the OGS_CONFIG must be changed, but also the OGS_LSOLVER (see https://github.com/ufz/ogs5/pull/18 ) ?

It is a trivial problem. But only after you solve it.

Another question: what is the latest supported version of PETSc by OGS 5.7.1?

Thanks.

Luca Urpi

+41446338372

Swiss Seismological Service

ETH Zürich, NO H 62

Sonneggstrasse 5

8092 Zürich

Norihiro Watanabe

unread,

May 23, 2017, 5:20:10 PM5/23/17

to ogs-...@googlegroups.com

Hi Luca,

Setting -DOGS_CONFIG=PESC in CMake should be enough. OGS_LSOLVER is
automatically set during cmake.

Regarding the PETSc version, I think current OGS5 supports till 3.5.
(Wenqing, please correct me if I'm wrong).

Best,
Nori

> --
> You received this message because you are subscribed to the Google
> Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs+u...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.

Luca Urpi

unread,

May 24, 2017, 5:41:06 AM5/24/17

to ogs-...@googlegroups.com

I tried again, at least on the machine I am using (Fedora 25, cmake 3.8.0), it does not set automatically with cmake or with cmake-gui (also, it does not work with MPI).

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-devs+unsubscribe@googlegroups.com <mailto:ogs-devs+unsubscribe@googlegroups.com>.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "ogs-devs" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-devs+unsubscribe@googlegroups.com.

渡邉教弘

unread,

May 24, 2017, 6:26:09 AM5/24/17

to ogs-...@googlegroups.com

i see. it's working on the test machine at ufz. i will check it later when i have time. thanks for reporting the problem.

n

From: ogs-...@googlegroups.com <ogs-...@googlegroups.com> on behalf of Luca Urpi <luca...@gmail.com>
Sent: Wednesday, May 24, 2017 6:41:05 PM
To: ogs-...@googlegroups.com
Subject: Re: OGS5 and PETSc

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-devs+u...@googlegroups.com.

Norihiro Watanabe

unread,

May 25, 2017, 6:12:04 PM5/25/17

to ogs-...@googlegroups.com

Hi,

sorry there was a typo. it should be -DOGS_CONFIG=PETC. I tested on my
machine with the following command and it is also working here.

cmake ../sources/ -DOGS_CONFIG=PETSC -DPESC_DIR=/opt/petsc/petsc-3.5.4/
-DPETSC_ARCH=gcc-release

maybe you need to remove the build dir once and test it again.

n

On 05/24/2017 06:41 PM, Luca Urpi wrote:
> I tried again, at least on the machine I am using (Fedora 25, cmake
> 3.8.0), it does not set automatically with cmake or with cmake-gui
> (also, it does not work with MPI).
>
>
> 2017-05-23 23:20 GMT+02:00 Norihiro Watanabe

> <norihiro...@aist.go.jp <mailto:norihiro...@aist.go.jp>>:

>
> Hi Luca,
>
> Setting -DOGS_CONFIG=PESC in CMake should be enough. OGS_LSOLVER is
> automatically set during cmake.
>
> Regarding the PETSc version, I think current OGS5 supports till 3.5.
> (Wenqing, please correct me if I'm wrong).
>
>
> Best,
> Nori
>
>
>
> On 05/22/2017 11:25 PM, Luca Urpi wrote:
>
> Dear OGS developers,
>
> although OGS5 is at the end of its development (or am I wrong?)
> may you update the page
>
> https://docs.opengeosys.org/docs/devguide5/getting-started/build-configuration#panel3-2
> <https://docs.opengeosys.org/docs/devguide5/getting-started/build-configuration#panel3-2>
>
> adding the information that to compile with PETSc not only the
> OGS_CONFIG must be changed, but also the OGS_LSOLVER (see
> https://github.com/ufz/ogs5/pull/18
> <https://github.com/ufz/ogs5/pull/18> ) ?
>
> It is a trivial problem. But only after you solve it.
>
> Another question: what is the latest supported version of PETSc
> by OGS 5.7.1?
>
> Thanks.
>
> Luca Urpi

> +41446338372 <tel:%2B41446338372>

>
> Swiss Seismological Service
> ETH Zürich, NO H 62
> Sonneggstrasse 5
> 8092 Zürich
>
>
> --
> You received this message because you are subscribed to the
> Google Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails from

> it, send an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>
> <mailto:ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>>.

> For more options, visit https://groups.google.com/d/optout

> <https://groups.google.com/d/optout>.

>
>
> --
> You received this message because you are subscribed to the Google
> Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails from it,

> send an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>.

> For more options, visit https://groups.google.com/d/optout

> <https://groups.google.com/d/optout>.

>
>
> --
> You received this message because you are subscribed to the Google
> Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails from it, send

> an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs+u...@googlegroups.com>.

Luca Urpi

unread,

May 26, 2017, 7:39:56 AM5/26/17

to ogs-...@googlegroups.com

Alright, yes the problem is with reusing the build dir.
If cmake is run for the first time, then it sets OGS_LSOLVER to the corresponding OGS_CONFIG: if no value is given it defaults to RF. Then if the same build folder is used afterwards, with different OGS_CONFIG, the OGS_LSOLVER needs to be set manually.

With cmake-gui, OGS_LSOLVER must be set manually, since it is set to RF with the first "configure".
However, if OGS_LSOLVER is set manually to DEFAULT, then it is updated automatically with the right value according to OGS_CONFIG, running "configure" for the second time.
Maybe there is an error and OGS_LSOLVER should be set to DEFAULT by the first "configure"? I am not very practical with cmake definitions and set-up.

Thanks for your patience.

Luca

2017-05-26 0:11 GMT+02:00 Norihiro Watanabe <norihiro...@aist.go.jp>:

Hi,

sorry there was a typo. it should be -DOGS_CONFIG=PETC. I tested on my machine with the following command and it is also working here.

cmake ../sources/ -DOGS_CONFIG=PETSC -DPESC_DIR=/opt/petsc/petsc-3.5.4/ -DPETSC_ARCH=gcc-release

maybe you need to remove the build dir once and test it again.

n

On 05/24/2017 06:41 PM, Luca Urpi wrote:

I tried again, at least on the machine I am using (Fedora 25, cmake 3.8.0), it does not set automatically with cmake or with cmake-gui (also, it does not work with MPI).

2017-05-23 23:20 GMT+02:00 Norihiro Watanabe <norihiro...@aist.go.jp <mailto:norihiro.watanabe@aist.go.jp>>:

it, send an email to ogs-devs+unsubscribe@googlegroups.com
<mailto:ogs-devs%2Bunsubscribe@googlegroups.com>
<mailto:ogs-devs+unsubscribe@googlegroups.com
<mailto:ogs-devs%2Bunsubscribe@googlegroups.com>>.

For more options, visit https://groups.google.com/d/optout
<https://groups.google.com/d/optout>.

-- You received this message because you are subscribed to the Google
Groups "ogs-devs" group.
To unsubscribe from this group and stop receiving emails from it,

send an email to ogs-devs+unsubscribe@googlegroups.com
<mailto:ogs-devs%2Bunsubscribe@googlegroups.com>.

For more options, visit https://groups.google.com/d/optout
<https://groups.google.com/d/optout>.

--
You received this message because you are subscribed to the Google Groups "ogs-devs" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-devs+unsubscribe@googlegroups.com <mailto:ogs-devs+unsubscribe@googlegroups.com>.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "ogs-devs" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-devs+unsubscribe@googlegroups.com.

Norihiro Watanabe

unread,

May 28, 2017, 6:28:03 PM5/28/17

to ogs-...@googlegroups.com

many thanks for reporting the details. The problem should be fixed in
#81 (https://github.com/ufz/ogs5/pull/81).

best,
nori

On 05/26/2017 08:39 PM, Luca Urpi wrote:
> Alright, yes the problem is with reusing the build dir.
> If cmake is run for the first time, then it sets OGS_LSOLVER to the
> corresponding OGS_CONFIG: if no value is given it defaults to RF. Then
> if the same build folder is used afterwards, with different OGS_CONFIG,
> the OGS_LSOLVER needs to be set manually.
>
> With cmake-gui, OGS_LSOLVER must be set manually, since it is set to RF
> with the first "configure".
> However, if OGS_LSOLVER is set manually to DEFAULT, then it is updated
> automatically with the right value according to OGS_CONFIG, running
> "configure" for the second time.
> Maybe there is an error and OGS_LSOLVER should be set to DEFAULT by the
> first "configure"? I am not very practical with cmake definitions and
> set-up.
>
> Thanks for your patience.
>
> Luca
>
> 2017-05-26 0:11 GMT+02:00 Norihiro Watanabe

> <norihiro...@aist.go.jp <mailto:norihiro...@aist.go.jp>>:

>
> Hi,
>
> sorry there was a typo. it should be -DOGS_CONFIG=PETC. I tested on
> my machine with the following command and it is also working here.
>
> cmake ../sources/ -DOGS_CONFIG=PETSC
> -DPESC_DIR=/opt/petsc/petsc-3.5.4/ -DPETSC_ARCH=gcc-release
>
> maybe you need to remove the build dir once and test it again.
>
> n
>
>
> On 05/24/2017 06:41 PM, Luca Urpi wrote:
>
> I tried again, at least on the machine I am using (Fedora 25,
> cmake 3.8.0), it does not set automatically with cmake or with
> cmake-gui (also, it does not work with MPI).
>
>
> 2017-05-23 23:20 GMT+02:00 Norihiro Watanabe
> <norihiro...@aist.go.jp

> <mailto:norihiro...@aist.go.jp>
> <mailto:norihiro...@aist.go.jp
> <mailto:norihiro...@aist.go.jp>>>:

> +41446338372 <tel:%2B41446338372> <tel:%2B41446338372>

>
> Swiss Seismological Service
> ETH Zürich, NO H 62
> Sonneggstrasse 5
> 8092 Zürich
>
>
> -- You received this message because you are
> subscribed to the
> Google Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving
> emails from
> it, send an email to

> ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>
> <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <mailto:ogs-devs%252Buns...@googlegroups.com>>
> <mailto:ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>
> <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <mailto:ogs-devs%252Buns...@googlegroups.com>>>.

> For more options, visit
> https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>.
>
>
> -- You received this message because you are subscribed
> to the Google
> Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails
> from it,

> send an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>
> <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <mailto:ogs-devs%252Buns...@googlegroups.com>>.

> For more options, visit https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>.
>
>
> --
> You received this message because you are subscribed to the
> Google Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails from

> it, send an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>
> <mailto:ogs-devs+u...@googlegroups.com

> <mailto:ogs-devs%2Bunsu...@googlegroups.com>>.

> For more options, visit https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>.
>
>
> --
> You received this message because you are subscribed to the Google
> Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails from it,

> send an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>.

> For more options, visit https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>.
>
>
> --
> You received this message because you are subscribed to the Google
> Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails from it, send

> an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs+u...@googlegroups.com>.

Chan-Hee Park

unread,

May 31, 2017, 1:45:30 AM5/31/17

to ogs-devs

Hello Nori,

I have followed your latest patch just merged on CentOS 7. The compilation works just fine with the configuration below. However, when I tried it with KueperProblem-PS benchmark, it works great with the following.

[nbeyond@cc KueperProblem-PS]$ mpirun -f mfile -np 3 $ogs_petsc kueper

My question from here is: How can I prepare kueper_partitioned_nodes_3.msh, kueper_partitioned_elems_3.msh, and kueper_partitioned_cfg3.msh? I guess those should come from

partmesh --ogs2metis kueper
mpmetis kueper.mesh
partmesh --metis2ogs -np 3 -n kueper

I do not see how to get the last one ( kueper_partitioned_cfg3.msh). I appreciate any help on this.

One more question, please. Would the latest ogs 5 with the patch work with HT problems on sequential mode if the configuration files are prepared just as I talked? This is extremely important because it is all about scale-up of applications to multi-million nodes. Thanks in advance.

Best regards,

> <mailto:ogs-devs%252Bunsubscribe@googlegroups.com>>

> <mailto:ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>
> <mailto:ogs-devs%2Bunsu...@googlegroups.com

> <mailto:ogs-devs%252Bunsubscribe@googlegroups.com>>>.

> For more options, visit
> https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>.
>
>
> -- You received this message because you are subscribed
> to the Google
> Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails
> from it,
> send an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs%2Bunsu...@googlegroups.com>
> <mailto:ogs-devs%2Bunsu...@googlegroups.com

> <mailto:ogs-devs%252Bunsubscribe@googlegroups.com>>.

Norihiro Watanabe

unread,

May 31, 2017, 1:54:41 AM5/31/17

to ogs-...@googlegroups.com

Hi,

On 05/31/2017 02:45 PM, Chan-Hee Park wrote:
> Hello Nori,
>
> I have followed your latest patch just merged on CentOS 7. The
> compilation works just fine with the configuration below. However, when
> I tried it with KueperProblem-PS benchmark, it works great with the
> following.
>
> [nbeyond@cc KueperProblem-PS]$ mpirun -f mfile -np 3 $ogs_petsc kueper
>
> My question from here is: How can I
> prepare kueper_partitioned_nodes_3.msh, kueper_partitioned_elems_3.msh,
> and kueper_partitioned_cfg3.msh? I guess those should come from
>
> partmesh --ogs2metis kueper
> mpmetis kueper.mesh
> partmesh --metis2ogs -np 3 -n kueper
>
>

> I do not see how to get the last one ( *kueper_partitioned_cfg3.msh*). I

> appreciate any help on this.

Have you tested with the latest partmesh available from
https://github.com/ufz/mesh_partition? I think Wenqing has made some
changes and it generates _cfg*.

> One more question, please. Would the latest ogs 5 with the patch work
> with HT problems on sequential mode if the configuration files are
> prepared just as I talked? This is extremely important because it is all
> about scale-up of applications to multi-million nodes. Thanks in advance.

What do you mean sequential mode? sequential coupling? HT processes
should also work with petsc.

n

>
> Best regards,
>
>
> On Monday, May 29, 2017 at 7:28:03 AM UTC+9, Norihiro Watanabe wrote:
>
> many thanks for reporting the details. The problem should be fixed in
> #81 (https://github.com/ufz/ogs5/pull/81

> <https://github.com/ufz/ogs5/pull/81>).

> > <norihiro...@aist.go.jp <javascript:>
> <mailto:norihiro...@aist.go.jp <javascript:>>>:

> >
> > Hi,
> >
> > sorry there was a typo. it should be -DOGS_CONFIG=PETC. I
> tested on
> > my machine with the following command and it is also working
> here.
> >
> > cmake ../sources/ -DOGS_CONFIG=PETSC
> > -DPESC_DIR=/opt/petsc/petsc-3.5.4/ -DPETSC_ARCH=gcc-release
> >
> > maybe you need to remove the build dir once and test it again.
> >
> > n
> >
> >
> > On 05/24/2017 06:41 PM, Luca Urpi wrote:
> >
> > I tried again, at least on the machine I am using (Fedora
> 25,
> > cmake 3.8.0), it does not set automatically with cmake or
> with
> > cmake-gui (also, it does not work with MPI).
> >
> >
> > 2017-05-23 23:20 GMT+02:00 Norihiro Watanabe

> > <norihiro...@aist.go.jp <javascript:>
> > <mailto:norihiro...@aist.go.jp <javascript:>>
> > <mailto:norihiro...@aist.go.jp <javascript:>
> > <mailto:norihiro...@aist.go.jp <javascript:>>>>:

> > ogs-devs+u...@googlegroups.com <javascript:>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>
> > <mailto:ogs-devs%252Buns...@googlegroups.com
> <javascript:>>>
> > <mailto:ogs-devs+u...@googlegroups.com
> <javascript:>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>
> > <mailto:ogs-devs%252Buns...@googlegroups.com
> <javascript:>>>>.

> > For more options, visit
> > https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>>.
> >
> >
> > -- You received this message because you are
> subscribed
> > to the Google
> > Groups "ogs-devs" group.
> > To unsubscribe from this group and stop receiving
> emails
> > from it,
> > send an email to ogs-devs+u...@googlegroups.com

> <javascript:>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>
> > <mailto:ogs-devs%252Buns...@googlegroups.com
> <javascript:>>>.

> > For more options, visit
> https://groups.google.com/d/optout <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>>.
> >
> >
> > --
> > You received this message because you are subscribed to the
> > Google Groups "ogs-devs" group.
> > To unsubscribe from this group and stop receiving emails
> from
> > it, send an email to ogs-devs+u...@googlegroups.com

> <javascript:>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>>
> > <mailto:ogs-devs+u...@googlegroups.com <javascript:>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>>>.

> > For more options, visit
> https://groups.google.com/d/optout <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>.
> >
> >
> > --
> > You received this message because you are subscribed to the
> Google
> > Groups "ogs-devs" group.
> > To unsubscribe from this group and stop receiving emails from
> it,

> > send an email to ogs-devs+u...@googlegroups.com <javascript:>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com <javascript:>>.

> > For more options, visit https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>.
> >
> >
> > --
> > You received this message because you are subscribed to the Google
> > Groups "ogs-devs" group.
> > To unsubscribe from this group and stop receiving emails from it,
> send

> > an email to ogs-devs+u...@googlegroups.com <javascript:>
> > <mailto:ogs-devs+u...@googlegroups.com <javascript:>>.

Chan-Hee Park

unread,

May 31, 2017, 4:46:33 AM5/31/17

to ogs-devs

Thanks, Nori. It works with the new mesh_partition given by WW.

Now, another problem with PVD in the .out file. It doesn't output .pvtu files to combine .vtu of each decomposed domain to have a whole domain. Does this need update Nori?

> > <mailto:ogs-devs%252Bunsubscribe@googlegroups.com

> <javascript:>>>
> > <mailto:ogs-devs+u...@googlegroups.com
> <javascript:>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>

> > <mailto:ogs-devs%252Bunsubscribe@googlegroups.com

> <javascript:>>>>.
> > For more options, visit
> > https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>>.
> >
> >
> > -- You received this message because you are
> subscribed
> > to the Google
> > Groups "ogs-devs" group.
> > To unsubscribe from this group and stop receiving
> emails
> > from it,
> > send an email to ogs-devs+u...@googlegroups.com
> <javascript:>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>>
> > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>

> > <mailto:ogs-devs%252Bunsubscribe@googlegroups.com

Norihiro Watanabe

unread,

May 31, 2017, 4:51:07 PM5/31/17

to ogs-...@googlegroups.com

I don't think PVTU output is supported in the official ogs5. you can use
Wenqing's script

https://github.com/wenqing/post_process_ogs5

On 31.05.17 17:46, Chan-Hee Park wrote:
> Thanks, Nori. It works with the new mesh_partition given by WW.
>
> Now, another problem with PVD in the .out file. It doesn't output

> .*pvtu* files to combine .*vtu* of each decomposed domain to have a

> whole domain. Does this need update Nori?
>
> On Wednesday, May 31, 2017 at 2:54:41 PM UTC+9, Norihiro Watanabe wrote:
>
> Hi,
>
> On 05/31/2017 02:45 PM, Chan-Hee Park wrote:
> > Hello Nori,
> >
> > I have followed your latest patch just merged on CentOS 7. The
> > compilation works just fine with the configuration below. However,
> when
> > I tried it with KueperProblem-PS benchmark, it works great with the
> > following.
> >
> > [nbeyond@cc KueperProblem-PS]$ mpirun -f mfile -np 3 $ogs_petsc
> kueper
> >
> > My question from here is: How can I
> > prepare kueper_partitioned_nodes_3.msh,
> kueper_partitioned_elems_3.msh,
> > and kueper_partitioned_cfg3.msh? I guess those should come from
> >
> > partmesh --ogs2metis kueper
> > mpmetis kueper.mesh
> > partmesh --metis2ogs -np 3 -n kueper
> >
> >
> > I do not see how to get the last one (
> *kueper_partitioned_cfg3.msh*). I
> > appreciate any help on this.
>
> Have you tested with the latest partmesh available from
> https://github.com/ufz/mesh_partition

> <https://github.com/ufz/mesh_partition>? I think Wenqing has made some

> > <javascript:>>
> > >
> <mailto:ogs-devs%2Bunsu...@googlegroups.com <javascript:>
> > <javascript:>
> > > <mailto:ogs-devs%252Buns...@googlegroups.com
> <javascript:>

> > <javascript:>>>
> > > <mailto:ogs-devs+u...@googlegroups.com
> > <javascript:>
> > > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>

> > <javascript:>>
> > >
> <mailto:ogs-devs%2Bunsu...@googlegroups.com <javascript:>
> > <javascript:>
> > > <mailto:ogs-devs%252Buns...@googlegroups.com
> <javascript:>

> > <javascript:>>
> > > <mailto:ogs-devs+u...@googlegroups.com <javascript:>
> > > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>

> > <javascript:>>>.
> > > For more options, visit
> > https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>
> > > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>>.
> > >
> > >
> > > --
> > > You received this message because you are subscribed to
> the
> > Google
> > > Groups "ogs-devs" group.
> > > To unsubscribe from this group and stop receiving
> emails from
> > it,
> > > send an email to ogs-devs+u...@googlegroups.com
> <javascript:>
> > > <mailto:ogs-devs%2Bunsu...@googlegroups.com

> <javascript:> <javascript:>>.

> > > For more options, visit
> https://groups.google.com/d/optout <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>
> > > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>>.
> > >
> > >
> > > --
> > > You received this message because you are subscribed to the
> Google
> > > Groups "ogs-devs" group.
> > > To unsubscribe from this group and stop receiving emails
> from it,
> > send
> > > an email to ogs-devs+u...@googlegroups.com <javascript:>

> > > <mailto:ogs-devs+u...@googlegroups.com <javascript:>>.
> > > For more options, visit https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>
> > <https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>>.
> >
> > --
> > You received this message because you are subscribed to the Google
> > Groups "ogs-devs" group.
> > To unsubscribe from this group and stop receiving emails from it,
> send
> > an email to ogs-devs+u...@googlegroups.com <javascript:>
> > <mailto:ogs-devs+u...@googlegroups.com <javascript:>>.
> > For more options, visit https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>.
>
> --
> You received this message because you are subscribed to the Google
> Groups "ogs-devs" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ogs-devs+u...@googlegroups.com
> <mailto:ogs-devs+u...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.

--
渡邉教弘（工博 Dr.-Ing.）

国立研究開発法人産業技術総合研究所
再生可能エネルギー研究所（FREA）
再生可能エネルギー研究センター
地熱チーム　研究員

〒963-0298 福島県郡山市待池台二丁目２番９
TEL: 029-861-4463 (内222-44463)
FAX: 024-963-0828
E-MAIL: norihiro...@aist.go.jp
URL: http://www.aist.go.jp/fukushima/

Chan-Hee Park

unread,

Jun 1, 2017, 1:01:10 AM6/1/17

to ogs-devs

I tried Wenqing's script. Didn't work with my example. Kueper-PcPnw-petsc. However, VTK outputs everything ok. Great work with petsc.

> > > <mailto:ogs-devs%252Bunsubscribe@googlegroups.com

> <javascript:>
> > <javascript:>>>
> > > <mailto:ogs-devs+u...@googlegroups.com
> > <javascript:>
> > > <mailto:ogs-devs%2Bunsu...@googlegroups.com
> <javascript:>
> > <javascript:>>
> > >
> <mailto:ogs-devs%2Bunsu...@googlegroups.com <javascript:>
> > <javascript:>

> > > <mailto:ogs-devs%252Bunsubscribe@googlegroups.com

Reply all

Reply to author

Forward