Dear Colleagues, I am trying to install ETK in hopper, at NERSC. I started from scratch following the steps of the advanced tutorial. I am building a package, which includes an N-body code and Cactus, so I my aim is to be able to run Cactus in different machines. My last attempt showed me following issue: After doing simfactory/bin/sim setup in hopper, I tried
simfactory/bin/sim build
with the result
Using configuration: sim Reconfiguring sim Writing configuration to: /global/u1/f/fiestas/CACTUS-save/Cactus/configs/sim/OptionList /bin/bash: module: command not found
I am using the last development version here http://einsteintoolkit.org/download/ (also tried the last release with the same results)
I found in simfactory/mdb/machines/hopper.ini a list of module commands which apparently can not be executed. I run it in queenbee without problems. Would appreciate any help.
Regards, Jose
On Sat, Feb 25, 2012 at 11:51 AM, Jose Fiestas Iquira jafiestas@lbl.govwrote:
Hi Erik,
Yes, should be loaded automatically. Find strange they don't. I use ./simfactory/bin/sim build from the Cactus directory (I tried adding --thornlist=manifest/einsteintoolkit.th) but I keep getting Using configuration: sim Reconfiguring sim Writing configuration to: /global/u1/f/fiestas/CACTUS/Cactus/configs/sim/OptionList /bin/bash: module: command not found
I tried deleting the module commands in simfactory/mdb/machines/hopper.ini and doing them by hand. I got more output (attached) but an error anyway.
This tries are from scratch, I loaded ETK from http://einsteintoolkit.org/download/ the development version, as you tried.
It seems to me the problem is related to the configuration of my account. But I did not change the defaults.
Best, Jose
On Sat, Feb 25, 2012 at 6:16 AM, Erik Schnetter schnetter@cct.lsu.eduwrote:
Jose
The modules that are loaded automatically should not matter; Simfactory should do the right thing anyway.
What command do you use to compile? I use
./bin/sim build --thornlist=manifest/einsteintoolkit.th
-erik
On Fri, Feb 24, 2012 at 6:36 PM, Jose Fiestas Iquira jafiestas@lbl.gov wrote:
Sorri pgi compilers, not pig :)
On Fri, Feb 24, 2012 at 3:35 PM, Jose Fiestas Iquira <jafiestas@lbl.gov
wrote:
You have loaded the pig compilers. I tried both pig and intel without success. It is something related to the configuration. I get: Internal error - couldn't enter '/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/config-data'
Cheers, Jose
On Fri, Feb 24, 2012 at 3:30 PM, Erik Schnetter <schnetter@cct.lsu.edu
wrote:
$ module list Currently Loaded Modulefiles:
- modules/3.2.6.6
- xtpe-network-gemini
- pgi/11.9.0
- xt-libsci/11.0.03
- udreg/2.3.1-1.0400.3911.5.13.gem
- ugni/2.3-1.0400.4127.5.20.gem
- pmi/3.0.0-1.0000.8661.28.2807.gem
- dmapp/3.2.1-1.0400.3965.10.63.gem
- gni-headers/2.1-1.0400.4156.6.1.gem
- xpmem/0.1-2.0400.30792.5.6.gem
- xe-sysroot/4.0.36
- xt-asyncpe/5.04
- atp/1.3.0
- PrgEnv-pgi/4.0.36
- eswrap/1.0.10
- xtpe-mc12
- xt-shmem/5.4.0
- xt-mpich2/5.4.0
- torque/2.5.9
- moab/6.1.4
-erik
On Fri, Feb 24, 2012 at 4:31 PM, Jose Fiestas Iquira <
jafiestas@lbl.gov>
wrote:
Erik, I suspect the default modules I loaded are not sufficient, because
I am
not doing more than 'simfactory/bin/sim setup' and 'simfactory/bin/sim build'. Could you please send me a 'module list' of your account to compare with mine? Best, Jose
On Fri, Feb 24, 2012 at 12:26 PM, Erik Schnetter schnetter@cct.lsu.edu wrote: > > Jose > > I just built the ET from scratch and ran the test cases there. All > went fine on Hopper, no problems at all. > > -erik > > On Fri, Feb 24, 2012 at 2:39 PM, Erik Schnetter > schnetter@cct.lsu.edu > wrote: > > On Fri, Feb 24, 2012 at 2:11 PM, Jose Fiestas Iquira > > jafiestas@lbl.gov > > wrote: > >> Hi Erik, > >> > >> I am getting in Hopper the same error after simfactory/bin/sim > >> build: > >> Using configuration: sim > >> Reconfiguring sim > >> Writing configuration to: > >> /global/u1/f/fiestas/CACTUS/Cactus/configs/sim/OptionList > >> /bin/bash: module: command not found > >> fiestas@hopper10:~/CACTUS/Cactus> exit > >> exit > >> > >> Could you install it without problems? > > > > Yes. I am not seeing these problems. Maybe they are specific to
your
> > account settings? Could you investigate where the "module"
command
> > is > > used? > > > >> A question. Should I set hopper10 as: > >> key allocation [] > > > > I don't know what you mean by this. > > > > The allocation on Hopper is the "repository" there; it is e.g. > > "m152" > > for me. > > > > -erik, > > > >> Thanks, > >> Jose > >> > >> > >> On Fri, Feb 24, 2012 at 10:45 AM, Jose Fiestas Iquira > >> jafiestas@lbl.gov > >> wrote: > >>> > >>> You mean this version? > >>> > >>>
http://svn.einsteintoolkit.org/manifest/trunk/einsteintoolkit.th
> >>> > >>> I am trying now in a new directory. I had installed Cactus, the > >>> Wavetoy > >>> and ETk in the same place and probably there were some
conflicts.
> >>> That > >>> is > >>> where I got the module error. Hope doing it new works well as
in
> >>> your > >>> case. > >>> Did you follow the tutorial in the ETK page for new users? > >>> > >>> Cheers, > >>> Jose > >>> > >>> > >>> > >>> > >>> On Fri, Feb 24, 2012 at 10:18 AM, Erik Schnetter > >>> schnetter@cct.lsu.edu > >>> wrote: > >>>> > >>>> Jose > >>>> > >>>> No, things didn't work on Carver. I received errors, but
didn't
> >>>> investigate, mostly because I dislike the PGI compiler. > >>>> > >>>> I'm building on Hopper as we speak, using the Intel compiler,
and
> >>>> things are going fine. > >>>> > >>>> I'm using the development version of the ET, not the release
from
> >>>> last > >>>> autumn, though. > >>>> > >>>> -erik > >>>> > >>>> On Fri, Feb 24, 2012 at 12:40 PM, Jose Fiestas Iquira > >>>> jafiestas@lbl.gov > >>>> wrote: > >>>> > Hi Erik, > >>>> > The problem with module (?) came actually when I tried it in > >>>> > Hopper. > >>>> > There > >>>> > is no special need for Carver, Hopper would be fine. Did
your
> >>>> > try > >>>> > in > >>>> > Carver > >>>> > worked, anyway? > >>>> > Best, > >>>> > Jose > >>>> > > >>>> > > >>>> > On Fri, Feb 24, 2012 at 9:14 AM, Erik Schnetter > >>>> > schnetter@cct.lsu.edu > >>>> > wrote: > >>>> >> > >>>> >> On Wed, Feb 22, 2012 at 8:38 PM, Jose Fiestas Iquira > >>>> >> jafiestas@lbl.gov > >>>> >> wrote: > >>>> >> > It was like these: > >>>> >> > .... > >>>> >> > COMPILING > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> >
/global/u1/f/fiestas/CACTUS/Cactus/arrangements/Carpet/LoopControl/src/loopcontrol.c
> >>>> >> > PGC-W-0155-parallel for loop variable must be local k > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> >
(/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/build/LoopControl/loopcontrol.c:
> >>>> >> > 1080) > >>>> >> > PGC-S-0094-Illegal type conversion required > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> >
(/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/build/LoopControl/loopcontrol.c:
> >>>> >> > 1080) > >>>> >> > PGC-S-0094-Illegal type conversion required > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> >
(/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/build/LoopControl/loopcontrol.c:
> >>>> >> > 1080) > >>>> >> > PGC/x86-64 Linux 9.0-3: compilation completed with severe > >>>> >> > errors > >>>> >> > make[3]: *** [loopcontrol.c.o] Error 2 > >>>> >> > make[2]: *** [make.checked] Error 2 > >>>> >> > make[1]: *** > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> >
[/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/lib/libthorn_LoopControl.a]
> >>>> >> > Error 2 > >>>> >> > make[1]: *** Waiting for unfinished jobs.... > >>>> >> > ..... > >>>> >> > COMPILING > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> >
/global/u1/f/fiestas/CACTUS/Cactus/arrangements/EinsteinInitialData/IDAxiBrillBH/src/ParamCheck.c
> >>>> >> > Creating > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> >
/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/lib/libthorn_IDAnalyticBH.a
> >>>> >> > Creating > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> >
/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/lib/libthorn_IDAxiBrillBH.a
> >>>> >> > Creating > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> > > >>>> >> >
/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/lib/libthorn_IDAxiOddBrillBH.a
> >>>> >> > make: *** [sim] Error 2 > >>>> >> > > >>>> >> > I am trying hopper now. Let see. > >>>> >> > >>>> >> Is there a reason why you want to use Carver, or would
another
> >>>> >> machine > >>>> >> be fine as well? > >>>> >> > >>>> >> The PGI compilers are often a bit peculiar; I prefer the
Intel
> >>>> >> compilers. Let me try to build on Hopper -- this should
work
> >>>> >> fine, > >>>> >> I > >>>> >> can then help you there. > >>>> >> > >>>> >> -erik > >>>> >> > >>>> >> -- > >>>> >> Erik Schnetter schnetter@cct.lsu.edu > >>>> >> http://www.perimeterinstitute.ca/personal/eschnetter/ > >>>> > > >>>> > > >>>> > >>>> > >>>> > >>>> -- > >>>> Erik Schnetter schnetter@cct.lsu.edu > >>>> http://www.perimeterinstitute.ca/personal/eschnetter/ > >>> > >>> > >> > > > > > > > > -- > > Erik Schnetter schnetter@cct.lsu.edu > > http://www.perimeterinstitute.ca/personal/eschnetter/ > > > > -- > Erik Schnetter schnetter@cct.lsu.edu > http://www.perimeterinstitute.ca/personal/eschnetter/
-- Erik Schnetter schnetter@cct.lsu.edu http://www.perimeterinstitute.ca/personal/eschnetter/
-- Erik Schnetter schnetter@cct.lsu.edu http://www.perimeterinstitute.ca/personal/eschnetter/
On Wed, Mar 07, 2012 at 05:33:51PM -0800, Jose Fiestas Iquira wrote:
I found in simfactory/mdb/machines/hopper.ini a list of module commands which apparently can not be executed. I run it in queenbee without problems. Would appreciate any help.
It would be interesting which of the modules commands fail. The software environment on hopper probably changed. Can you try to execute them by hand? Maybe all that is needed to do is to specify some new version.
Frank
Actually it can not execute the module command itself Inside hopper.ini should be executing
envsetup = module unload PrgEnv-cray PrgEnv-gnu PrgEnv-intel PrgEnv-pathscale PrgEnv-pgi && module load PrgEnv-intel && module load acml
but it does not come to it. I tried it by hand and commented it. But I got the following error:
................. End of options from '/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/OptionList'. Adding configuration options from user defaults in /global/homes/f/fiestas/.cactus/config... End of options from user defaults. Internal error - couldn't enter '/global/u1/f/fiestas/CACTUS/Cactus/configs/sim/config-data'
Error reconfiguring sim-config make: *** [sim-config] Error 2 ..................
I am now trying to use gmake and create my own config file. But want only to use the maclachnan thorn and its dependencies. So, I need to create an own thornlist. By doing this and using ensteintoolkit.th or cactus...th or WaveDemo.th I obtain some errors from certain thorns, like BLAS, GIT, LORENE, HDF%, and few others.
Thanks, Jose
2012/3/8 Frank Loeffler knarf@cct.lsu.edu
On Wed, Mar 07, 2012 at 05:33:51PM -0800, Jose Fiestas Iquira wrote:
I found in simfactory/mdb/machines/hopper.ini a list of module commands which apparently can not be executed. I run it in queenbee without problems. Would appreciate any help.
It would be interesting which of the modules commands fail. The software environment on hopper probably changed. Can you try to execute them by hand? Maybe all that is needed to do is to specify some new version.
Frank
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux)
iQIcBAEBCAAGBQJPWMS7AAoJEOkzpip+I59k4GIP/17dnABLGEXUAkM5zMXonRrb pYXNX5qg05YGVGPYYyjiW10zWkvGjeUMcqxaPBmfJO5gRv2VbbNoD/ONf96XCUI0 qypArGlu+KvYvFZOFYdHJjEV54wVVRs0umh8CJh76+E2JtSgACUyJwsmGMWKVT/a GSHrl1iLinT75iGBB/5Pl5bbI8P8WVqjpm1I0TsPyftsdXUrtH63ejdaxdjW4AQ4 rMKKnDcGW6sffM4bHAHU6sPK5drmwvB9hbOq77m8PWlywrZkptN1CSAE1VrFsmDX +mA4CvMTLARkiZCqv0kYxfA0c3hMgiqc6ShTBA8Hp5J09vVI1PI2TRWh2//dVLE/ VhwltdUWVGfMZXwmOb5r6u0VJ4y6H4T2bWE7m3dveWqxvd05D53BS3389RcD2Nrn iMc5Iq7L8wkV/WDYQg2mylYeESNXJ2RrCRe1MDIibO0KP26rxEmvFhV4pd/irrw3 /yyDUsGQljHoNthewDPuTLkhw5CkQEgVi/Ebg68VIgMDVnt+hIUgbYW7kgo8SMMx ZnTbOW5pGJ2skdpFtvRLG2ZvzK6I0dPTzq0bNb1bWyGrdrStPnhzBW9SEjDK/ca4 IaKXmsKt+0j27C7FBBtZq7PvOcQmhm9OK1PhGNMV4MPnCI+Cbnr5cDoXpSC4CyxW pd4M/FXo12lKZhRUVtly =a6gN -----END PGP SIGNATURE-----
users@lists.einsteintoolkit.org