From mdas at umich.edu Sun Mar 2 23:46:32 2008 From: mdas at umich.edu (Mousumi Das) Date: Mon, 3 Mar 2008 00:46:32 -0500 (EST) Subject: [FLASH-USERS] 3D resolution In-Reply-To: <40347.75.3.130.101.1198859248.squirrel@flash.uchicago.edu> References: <475EB841.1080909@astro.umass.edu> <39762.75.3.145.172.1197899760.squirrel@flash.uchicago.edu> <55000.75.3.145.172.1197912594.squirrel@flash.uchicago.edu> <40347.75.3.130.101.1198859248.squirrel@flash.uchicago.edu> Message-ID: Hello All, How can one increase resolution in 3D FLASH? I am using nxb=nyb=nz=16 zones with refinement level 6 for box length 0 to 0.01cm. IDL plot gives unifrom grid size 0.195 micron. But I want grid size to be about 0.08micron. Is there any way in FLASH that one can start with small domain with high resolution and then increase the simulation box. Or is there any way that we can modify the checkpoint file from high resolution of small domain calculation and coarsening the data, modify the checkpoint file and feed them back to a large domain 3D run. thanks, Mousumi On Fri, 28 Dec 2007, Anshu Dubey wrote: > Try switching the IO mode between hdf5_serial and hdf5_parallel and see > if you do any better. > >> Hello All, >> Even if I increase the frequency of the checkpointing, that is not >> going to help, bacause the code stops running while writting the 2nd >> checkpoint file always, chk_0002. Log file shows the follwoing error mesg. >> >> [ 12-28-2007 01:45.36 ] step: n=163 t=1.969343E-09 dt=8.117542E-12 >> [ 12-28-2007 01:47.06 ] step: n=164 t=1.985578E-09 dt=8.124520E-12 >> [ 12-28-2007 01:48.50 ] file_wr_open: type=checkpoint >> name=Ni_Tem_hdf5_chk_0002 >> >> If I see the output file it gives the error mesg >> [CHECKPOINT_WR] Writing checkpoint file Ni_Tem_hdf5_chk_0002 >> Progress: |......... >> Signal:11 info.si_errno:0(Success) si_code:1(SEGV_MAPERR) >> Failing at addr:0x2abd18f1e8 >> *** End of error message *** >> >> I have checked there is enough space for wrting files. Did anyone of you >> face such problem before? >> thanks, >> Mousumi >> >> >> On Mon, 17 Dec 2007, Anshu Dubey wrote: >> >>> The checkpoint and plotfile numbers are not related to each other. >>> For the plotfile, the number only provides the number from where >>> the plotfile count starts upon restart. >>> >>> But if your second checkpoint didn't get written correctly, I am afraid >>> you will have to restart from the first checkpoint, and repeat all the >>> steps in between. You might want to increase the frequence of >>> checkpointing if you are running into IO problems, that way you won't >>> have to rollback as much. >>> >>> Anshu >>>> >>>> I am using FLASH2.5. My programme stops while writing a checkpint >>>> file >>>> after outputing several plot files, but it has not completed the run >>>> till >>>> tstop value. >>>> >>>> [ 12-14-2007 21:45.22 ] step: n=364 t=5.630710E-09 dt=1.014811E-11 >>>> [ 12-14-2007 21:48.17 ] [AMR_REFINE_DEREFINE]: refinement initiated >>>> at >>>> 21:48.1 7 >>>> [ 12-14-2007 21:49.17 ] [AMR_REFINE_DEREFINE] blocks all: >>>> min=1326 >>>> max=1336 tot=13321 >>>> [ 12-14-2007 21:49.17 ] [AMR_REFINE_DEREFINE] blocks valid: >>>> min=1163 >>>> max=1167 tot=11656 >>>> [ 12-14-2007 21:49.17 ] [AMR_REFINE_DEREFINE]: refinement complete >>>> [ 12-14-2007 21:49.32 ] file_wr_open: type=checkpoint >>>> name=Ni_Tem_hdf5_chk_000 2 >>>> >>>> >>>> I want to restart the simulation with previous checkpointpoint file >>>> Ni_Tem_hdf5_chk_0001. According to FLASH manual I change the restart >>>> logical variable to .true. >>>> cpunumber I specfied 0001 ( last written checkpointfile) >>>> pltnumber 0001 ( last written plotfile number after 0001 check point) >>>> >>>> Although the last written pltnumber is 0056. But there is gap between >>>> checkpoint file chk_0001 and chk_0002. >>>> >>>> But with this I am not able to restart the simulation. How do I restart >>>> the simulation run if I want to get the plotfile (0057) after the last >>>> plotfile generated (0056). >>>> thanks, >>>> Mousumi >>>> >>>> On Mon, 17 Dec 2007, Anshu Dubey wrote: >>>> >>>>> People have used the parallel hdf5 module in Flash2.5 at the center >>>>> and still do, but on many platforms the restart is extremely slow. >>>>> We haven't been able to determine the cause. However, fortunately, >>>>> the problem didn't get carried over to Flash3, so we strongly >>>>> recommend switching over to Flash3 if it is possible for you. >>>>> >>>>> Anshu >>>>> >>>>>> Hi all, >>>>>> >>>>>> I have a serious problem when using the parallel hdf5 module to >>>>>> restart from a checkpoint file. It takes about half an hour to read a >>>>>> single variable to each node for a 1GB file. The same problem was >>>>>> reported two year ago >>>>>> (http://flash.uchicago.edu/pipermail/flash-users/2005-April/001938.html) >>>>>> but I cannot find the follow-ups. Did anyone succeed in using the >>>>>> parallel hdf5 module so far? Will the Flash center continue support >>>>>> such >>>>>> issue on Flash2.5? Thanks! >>>>>> >>>>>> Bests, >>>>>> Shikui >>>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>> >>> >>> >>> >> > > > > From smc at astro.as.utexas.edu Mon Mar 3 08:17:43 2008 From: smc at astro.as.utexas.edu (Sean Couch) Date: Mon, 3 Mar 2008 08:17:43 -0600 Subject: [FLASH-USERS] 3D resolution In-Reply-To: References: <475EB841.1080909@astro.umass.edu> <39762.75.3.145.172.1197899760.squirrel@flash.uchicago.edu> <55000.75.3.145.172.1197912594.squirrel@flash.uchicago.edu> <40347.75.3.130.101.1198859248.squirrel@flash.uchicago.edu> Message-ID: <224220EC-59AF-4FB6-9FC1-812D0A54CF8F@astro.as.utexas.edu> Hi Mousumi, Yes, there is a way. In flash.par, specify Nblock(x,y,z) = (some number > 1). This increases the number of blocks in the domain at refinement level one. nxb,nyb,nzb set the number of zones per block. You can also go beyond 6 levels of refinement. Sean --------------------------------------------- Sean M. Couch Graduate Research Assistant Department of Astronomy The University of Texas at Austin 1 University Station C1400 RLM 15.202A (512) 471-8316 smc at astro.as.utexas.edu www.as.utexas.edu/~smc On Mar 2, 2008, at 11:46 PM, Mousumi Das wrote: > Hello All, > How can one increase resolution in 3D FLASH? I am using > nxb=nyb=nz=16 zones with refinement level 6 for box length 0 to > 0.01cm. IDL plot gives unifrom grid size 0.195 micron. But I want > grid size to be about 0.08micron. Is there any way in FLASH that one > can start with small domain with high resolution and then increase > the simulation box. Or is there any way that we can modify the > checkpoint file from high resolution of small domain calculation and > coarsening the data, modify the checkpoint file and feed them back > to a large domain 3D run. thanks, > > Mousumi > > On Fri, 28 Dec 2007, Anshu Dubey wrote: > >> Try switching the IO mode between hdf5_serial and hdf5_parallel and >> see >> if you do any better. >> >>> Hello All, >>> Even if I increase the frequency of the checkpointing, that is >>> not >>> going to help, bacause the code stops running while writting the 2nd >>> checkpoint file always, chk_0002. Log file shows the follwoing >>> error mesg. >>> >>> [ 12-28-2007 01:45.36 ] step: n=163 t=1.969343E-09 dt=8.117542E-12 >>> [ 12-28-2007 01:47.06 ] step: n=164 t=1.985578E-09 dt=8.124520E-12 >>> [ 12-28-2007 01:48.50 ] file_wr_open: type=checkpoint >>> name=Ni_Tem_hdf5_chk_0002 >>> >>> If I see the output file it gives the error mesg >>> [CHECKPOINT_WR] Writing checkpoint file Ni_Tem_hdf5_chk_0002 >>> Progress: |......... >>> Signal:11 info.si_errno:0(Success) si_code:1(SEGV_MAPERR) >>> Failing at addr:0x2abd18f1e8 >>> *** End of error message *** >>> >>> I have checked there is enough space for wrting files. Did anyone >>> of you >>> face such problem before? >>> thanks, >>> Mousumi >>> >>> >>> On Mon, 17 Dec 2007, Anshu Dubey wrote: >>> >>>> The checkpoint and plotfile numbers are not related to each other. >>>> For the plotfile, the number only provides the number from where >>>> the plotfile count starts upon restart. >>>> >>>> But if your second checkpoint didn't get written correctly, I am >>>> afraid >>>> you will have to restart from the first checkpoint, and repeat >>>> all the >>>> steps in between. You might want to increase the frequence of >>>> checkpointing if you are running into IO problems, that way you >>>> won't >>>> have to rollback as much. >>>> >>>> Anshu >>>>> >>>>> I am using FLASH2.5. My programme stops while writing a >>>>> checkpint >>>>> file >>>>> after outputing several plot files, but it has not completed the >>>>> run >>>>> till >>>>> tstop value. >>>>> >>>>> [ 12-14-2007 21:45.22 ] step: n=364 t=5.630710E-09 >>>>> dt=1.014811E-11 >>>>> [ 12-14-2007 21:48.17 ] [AMR_REFINE_DEREFINE]: refinement >>>>> initiated >>>>> at >>>>> 21:48.1 7 >>>>> [ 12-14-2007 21:49.17 ] [AMR_REFINE_DEREFINE] blocks all: >>>>> min=1326 >>>>> max=1336 tot=13321 >>>>> [ 12-14-2007 21:49.17 ] [AMR_REFINE_DEREFINE] blocks valid: >>>>> min=1163 >>>>> max=1167 tot=11656 >>>>> [ 12-14-2007 21:49.17 ] [AMR_REFINE_DEREFINE]: refinement >>>>> complete >>>>> [ 12-14-2007 21:49.32 ] file_wr_open: type=checkpoint >>>>> name=Ni_Tem_hdf5_chk_000 2 >>>>> >>>>> >>>>> I want to restart the simulation with previous checkpointpoint >>>>> file >>>>> Ni_Tem_hdf5_chk_0001. According to FLASH manual I change the >>>>> restart >>>>> logical variable to .true. >>>>> cpunumber I specfied 0001 ( last written checkpointfile) >>>>> pltnumber 0001 ( last written plotfile number after 0001 check >>>>> point) >>>>> >>>>> Although the last written pltnumber is 0056. But there is gap >>>>> between >>>>> checkpoint file chk_0001 and chk_0002. >>>>> >>>>> But with this I am not able to restart the simulation. How do I >>>>> restart >>>>> the simulation run if I want to get the plotfile (0057) after >>>>> the last >>>>> plotfile generated (0056). >>>>> thanks, >>>>> Mousumi >>>>> >>>>> On Mon, 17 Dec 2007, Anshu Dubey wrote: >>>>> >>>>>> People have used the parallel hdf5 module in Flash2.5 at the >>>>>> center >>>>>> and still do, but on many platforms the restart is extremely >>>>>> slow. >>>>>> We haven't been able to determine the cause. However, >>>>>> fortunately, >>>>>> the problem didn't get carried over to Flash3, so we strongly >>>>>> recommend switching over to Flash3 if it is possible for you. >>>>>> >>>>>> Anshu >>>>>> >>>>>>> Hi all, >>>>>>> >>>>>>> I have a serious problem when using the parallel hdf5 module to >>>>>>> restart from a checkpoint file. It takes about half an hour to >>>>>>> read a >>>>>>> single variable to each node for a 1GB file. The same problem >>>>>>> was >>>>>>> reported two year ago >>>>>>> (http://flash.uchicago.edu/pipermail/flash-users/2005-April/001938.html >>>>>>> ) >>>>>>> but I cannot find the follow-ups. Did anyone succeed in using >>>>>>> the >>>>>>> parallel hdf5 module so far? Will the Flash center continue >>>>>>> support >>>>>>> such >>>>>>> issue on Flash2.5? Thanks! >>>>>>> >>>>>>> Bests, >>>>>>> Shikui >>>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>> >>>> >>>> >>>> >>> >> >> >> >> From turcotte at apollo.ubishops.ca Tue Mar 4 10:48:10 2008 From: turcotte at apollo.ubishops.ca (turcotte) Date: Tue, 4 Mar 2008 11:48:10 -0500 (EST) Subject: [FLASH-USERS] lrefine and parallel problem in flash 3.0 beta Message-ID: <200803041648.m24GmAO05535@apollo.ubishops.ca> Hello all, Using flash3.0 beta, if I run a job using lrefine_max =3 and then try running in parallel (on 4 cpus) the job always hang on the refinement step; last output in flash.log is ============================================================================== [GRID amr_refine_derefine] initiating refinement [GRID amr_refine_derefine] min blks 75 max blks 75 tot blks 600 [GRID amr_refine_derefine] min leaf blks 75 max leaf blks 75 tot leaf blk s 600 [GRID amr_refine_derefine] refinement complete [ 03-03-2008 20:55.58 ] [GRID gr_expandDomain]: iteration=1, create level=2 [GRID amr_refine_derefine] initiating refinement If I use lrefine_max =1 then I can run on multiple processors without problem. If I use lrefine_max =3 but run on a single processor, the job runs fine. The MPI version we have on the machine is MPICH version 1.2.6. Similar jobs run without any problem in flash2.5 with arbitrary lrefine and number of processors. Any ideas? Thanks, Sylvain --- Sylvain Turcotte Physics Department turcotte at apollo.ubishops.ca Bishop's University tel: (819) 822-9600 x 2287 Lennoxville (Quebec) fax: (819) 822-9661 Canada J1M 1Z7 office: J 11B From klaus at flash.uchicago.edu Wed Mar 5 11:58:32 2008 From: klaus at flash.uchicago.edu (Klaus Weide) Date: Wed, 5 Mar 2008 11:58:32 -0600 (CST) Subject: [FLASH-USERS] lrefine and parallel problem in flash 3.0 beta In-Reply-To: <200803041648.m24GmAO05535@apollo.ubishops.ca> References: <200803041648.m24GmAO05535@apollo.ubishops.ca> Message-ID: On Tue, 4 Mar 2008, turcotte wrote: > Using flash3.0 beta, if I run a job using lrefine_max =3 and then try running in > parallel (on 4 cpus) the job always hang on the refinement step; last output in > flash.log is > > ============================================================================== > [GRID amr_refine_derefine] initiating refinement > [GRID amr_refine_derefine] min blks 75 max blks 75 tot blks 600 > [GRID amr_refine_derefine] min leaf blks 75 max leaf blks 75 tot leaf blk > s 600 > [GRID amr_refine_derefine] refinement complete > [ 03-03-2008 20:55.58 ] [GRID gr_expandDomain]: iteration=1, create level=2 > [GRID amr_refine_derefine] initiating refinement Hi Sylvain, It seems the above is from a run on 8 cpus, since 600 (tot blks) = 75 (min/max blks) * 8. You seem to be starting with Nblockx*Nblocky*Nblockz = 75 root blocks, do you need such a high number of initial blocks? > If I use lrefine_max =1 then I can run on multiple processors without problem. > > If I use lrefine_max =3 but run on a single processor, the job runs fine. > > The MPI version we have on the machine is MPICH version 1.2.6. > > Similar jobs run without any problem in flash2.5 with arbitrary lrefine and > number of processors. > > Any ideas? Does the same problem occur with the FLASH3.0 code released last week? Does the hang still occur when you increase MAXBLOCKS? Can you reproduce it with fewer processors? With fewer root blocks? Also, try changing lrefine_min in addition to lrefine_max (to 2 or 3). Can you find out which process hangs where but using a debugger? Just some ideas, as you requested. Klaus From geoff at mso.anu.edu.au Mon Mar 10 19:31:59 2008 From: geoff at mso.anu.edu.au (Geoff Bicknell) Date: Tue, 11 Mar 2008 11:31:59 +1100 Subject: [FLASH-USERS] gmake of FLASH3.0 under Solaris 5 Message-ID: I have been having problems gmaking a FLASH executable under Solaris 5.10. The computer is a SUN 4 dual cpu machine with AMD Opteron processors. After a number of F90 modules have been successfully compiled I get the message: if [ -s "nameValueLL_data.mod" -a "nameValueLL_data.mod" -nt "namevaluell_data.mod" ];then ln -f nameValueLL_data.mod namevaluell_data.mod;fi /bin/sh: test: unknown operator -nt The problem seems to be that in the Bourne shell in Solaris 5.10 the test program does not recognise the -nt option. I have tried replacing /bin/sh in all the FLASH3.0 scripts by /bin/bash using the following command (issued from the top FLASH3.0 directory): find . -type f -exec fgres "/bin/sh" "/bin/bash" '{}' \; However, I still get the same error message. Any useful ideas would be appreciated. Thanks, Geoff Bicknell -- ================================================================================ Professor Geoffrey Bicknell, Associate Director (Academic) Research School of Astronomy & Astrophysics, Australian National University, Mt Stromlo Observatory, Cotter Rd., Weston ACT 2611, AUSTRALIA T:+61 (0)2 6125 0245 M: +61 (0)402 302 802 F: +61 (0)2 6125 0260 W: http://www.mso.anu.edu.au/~geoff ================================================================================ (ANU CRICOS #00120C) From carlo at oddjob.uchicago.edu Mon Mar 10 22:02:15 2008 From: carlo at oddjob.uchicago.edu (Carlo Graziani) Date: Mon, 10 Mar 2008 22:02:15 -0500 Subject: [FLASH-USERS] gmake of FLASH3.0 under Solaris 5 In-Reply-To: References: Message-ID: <47D5F637.7080000@oddjob.uchicago.edu> Geoff Bicknell wrote: > I have been having problems gmaking a FLASH executable under Solaris > 5.10. The computer is a SUN 4 dual cpu machine with AMD Opteron > processors. > > After a number of F90 modules have been successfully compiled I get > the message: > > if [ -s "nameValueLL_data.mod" -a "nameValueLL_data.mod" -nt > "namevaluell_data.mod" ];then ln -f nameValueLL_data.mod > namevaluell_data.mod;fi > /bin/sh: test: unknown operator -nt > > The problem seems to be that in the Bourne shell in Solaris 5.10 the > test program does not recognise the -nt option. > > I have tried replacing /bin/sh in all the FLASH3.0 scripts by > /bin/bash using the following command (issued from the top FLASH3.0 > directory): > > find . -type f -exec fgres "/bin/sh" "/bin/bash" '{}' \; > > However, I still get the same error message. > > Any useful ideas would be appreciated. > Hi Geoff. I believe that call to the shell is being generated by a system call in a Python script, and not directly from a Bourne shell script. So I don't think that tree-global search-and-replace can possibly do any good. Another possible strategy is to make a bin directory in your home directory, create a symlink from sh to /bin/bash (or to wherever your bash lives) in that directory, and make sure that that personal bin directory is first in your path when you start the build. That is, something like: mkdir $HOME/bin cd $HOME/bin ln -s /bin/bash sh cd env PATH=$HOME/bin:$PATH env PATH=$HOME/bin:$PATH make I don't have a solaris box to test on, so I can't say for sure that this will work, but it seems easy enough to try. A shell alias that lets bash impersonate sh might work too. Cheers, Carlo -- Carlo Graziani (773) 702-7973 (Voice) Department of Astronomy and Astrophysics (773) 702-6645 (FAX) University of Chicago 5640 South Ellis Avenue | If the free market really allocates resources Chicago, IL 60637 | efficiently, why does LA have four times as carlo at oddjob.uchicago.edu | many plastic surgeons as brain surgeons? From klaus at flash.uchicago.edu Tue Mar 11 09:46:15 2008 From: klaus at flash.uchicago.edu (Klaus Weide) Date: Tue, 11 Mar 2008 09:46:15 -0500 (CDT) Subject: [FLASH-USERS] gmake of FLASH3.0 under Solaris 5 In-Reply-To: References: Message-ID: On Tue, 11 Mar 2008, Geoff Bicknell wrote: > I have been having problems gmaking a FLASH executable under Solaris 5.10. > The computer is a SUN 4 dual cpu machine with AMD Opteron processors. > > After a number of F90 modules have been successfully compiled I get the > message: > > if [ -s "nameValueLL_data.mod" -a "nameValueLL_data.mod" -nt > "namevaluell_data.mod" ];then ln -f nameValueLL_data.mod > namevaluell_data.mod;fi > /bin/sh: test: unknown operator -nt > > The problem seems to be that in the Bourne shell in Solaris 5.10 the test > program does not recognise the -nt option. This comes from Makefile.Depends (automatically generated by setup_depends.py), which contains lines like these: namevaluell_data.mod: nameValueLL_data.mod ;if [ -s "$<" -a "$<" -nt "$@" ];then ln -f $< $@;fi (This is part of some trickery that tries to ensure that no more and no fewer than the necessary files are regenerated when gmake is run, including .mod files.) The simplest way to force gmake to use a certain shell is putting a line like SHELL=/bin/bash in a Makefile. The best place for this would probably be in sites/your.sitename/Makefile.h . Klaus Weide FLASH Center From Daniel.Pfenniger at obs.unige.ch Tue Mar 11 15:30:08 2008 From: Daniel.Pfenniger at obs.unige.ch (Daniel Pfenniger) Date: Tue, 11 Mar 2008 21:30:08 +0100 Subject: [FLASH-USERS] FLASH 3 hdf5 file format and Visit Message-ID: <47D6EBD0.7000008@obs.unige.ch> Hello, Has anyone also problem while reading the new hdf5 format in VisIt? I use VisIt 1.8.0 on Opteron RedHat 5.1 computers and both for the old and the new plugin the Visit message after trying to read hdf5 files is: --------------------- Error! VisIt was unable to open "/home/user1/FLASH3.0/model1_hdf5_chk_* database". Each attempt to open it caused VisIt's metadata server to crash. This can occur when the file is corrupted, or when the underlying file format has changed and VisIt's readers have not been updated yet, or when the reader VisIt is using for your file format is not robust. Please check whether the file is corrupted and, if not, contact a VisIt developer. ------------------------------ Daniel From hudson at mcs.anl.gov Tue Mar 11 15:46:14 2008 From: hudson at mcs.anl.gov (Randy Hudson) Date: Tue, 11 Mar 2008 15:46:14 -0500 Subject: [FLASH-USERS] FLASH 3 hdf5 file format and Visit In-Reply-To: <47D6EBD0.7000008@obs.unige.ch> References: <47D6EBD0.7000008@obs.unige.ch> Message-ID: <66EE4673-B883-49B6-884D-C70060267403@mcs.anl.gov> Daniel, Do you know if you already had a "~/.visit" directory before you built the new plugin? If so, did it contain a "config" file? If so, either move or delete it and restart visit. On Mar 11, 2008, at 3:30 PM, Daniel Pfenniger wrote: > Hello, > > Has anyone also problem while reading the new hdf5 format in VisIt? > I use VisIt 1.8.0 on Opteron RedHat 5.1 computers and both for the > old and the new plugin the Visit message after trying to read hdf5 > files is: > > --------------------- > Error! > > VisIt was unable to open "/home/user1/FLASH3.0/model1_hdf5_chk_* > database". Each attempt to open it caused VisIt's metadata server > to crash. This can occur when the file is corrupted, or when the > underlying file format has changed and VisIt's readers have not > been updated yet, or when the reader VisIt is using for your file > format is not robust. Please check whether the file is corrupted > and, if not, contact a VisIt developer. > > ------------------------------ > > > > Daniel > Randy. From seyit at astro.rug.nl Wed Mar 12 05:08:12 2008 From: seyit at astro.rug.nl (Seyit Hocuk) Date: Wed, 12 Mar 2008 11:08:12 +0100 Subject: [FLASH-USERS] FLASH 3 hdf5 file format and Visit In-Reply-To: <66EE4673-B883-49B6-884D-C70060267403@mcs.anl.gov> References: <47D6EBD0.7000008@obs.unige.ch> <66EE4673-B883-49B6-884D-C70060267403@mcs.anl.gov> Message-ID: <47D7AB8C.5050705@astro.rug.nl> I use Visit1.8.0 and have never had this error but I don't use the latest hdf5 format though. You are sure you don't fotrget to start visit with default format flash? Grtz Seyit. Randy Hudson wrote: > > Daniel, > > Do you know if you already had a "~/.visit" directory before you built > the new plugin? > > If so, did it contain a "config" file? > > If so, either move or delete it and restart visit. > > > On Mar 11, 2008, at 3:30 PM, Daniel Pfenniger wrote: > >> Hello, >> >> Has anyone also problem while reading the new hdf5 format in VisIt? >> I use VisIt 1.8.0 on Opteron RedHat 5.1 computers and both for the >> old and the new plugin the Visit message after trying to read hdf5 >> files is: >> >> --------------------- >> Error! >> >> VisIt was unable to open "/home/user1/FLASH3.0/model1_hdf5_chk_* >> database". Each attempt to open it caused VisIt's metadata server to >> crash. This can occur when the file is corrupted, or when the >> underlying file format has changed and VisIt's readers have not been >> updated yet, or when the reader VisIt is using for your file format >> is not robust. Please check whether the file is corrupted and, if >> not, contact a VisIt developer. >> >> ------------------------------ >> >> >> >> Daniel >> > > Randy. > > From seyit at astro.rug.nl Thu Mar 20 09:06:00 2008 From: seyit at astro.rug.nl (Seyit Hocuk) Date: Thu, 20 Mar 2008 15:06:00 +0100 Subject: [FLASH-USERS] Maximum number of blocks Message-ID: <47E26F48.7000205@astro.rug.nl> Hi all, I have a question about the maximum amount of blocks you can use for a simulation. I want to increase my resolution and give a high number for maxblocks, but I get an error that a common block or variable may not exceed 2147483647 bytes. See the error below. The setup is successfull "./setup jeans_SH -auto -maxblocks=90000", but while making it has this error. Is there a way to increase my resolution without this error? Thanks, Seyit cd object; make make[1]: Entering directory `/net/perenago/data/users/seyit/programmas/SimulationCodes/FLASH2.5/object' mpif90 -c -r8 -i4 -fast -ipo -ipo_obj -I /usr/local/mpich2/include -DN_DIM=2 -DMAXBLOCKS=90000 -DNXB=8 -DNYB=8 -DNZB=1 CosmologicalFunctions.F90 mpif90 -c -r8 -i4 -fast -ipo -ipo_obj -I /usr/local/mpich2/include -DN_DIM=2 -DMAXBLOCKS=90000 -DNXB=8 -DNYB=8 -DNZB=1 Friedmann.F90 mpif90 -c -r8 -i4 -fast -ipo -ipo_obj -I /usr/local/mpich2/include -DN_DIM=2 -DMAXBLOCKS=90000 -DNXB=8 -DNYB=8 -DNZB=1 physicaldata.F90 fortcom: Error: physicaldata.F90, line 196: A common block or variable may not exceed 2147483647 bytes kl_bnd:ku_bnd,maxblocks), TARGET :: unk .........................................^ compilation aborted for physicaldata.F90 (code 1) make[1]: *** [physicaldata.o] Error 1 From brockp at umich.edu Thu Mar 20 09:24:26 2008 From: brockp at umich.edu (Brock Palen) Date: Thu, 20 Mar 2008 10:24:26 -0400 Subject: [FLASH-USERS] Maximum number of blocks In-Reply-To: <47E26F48.7000205@astro.rug.nl> References: <47E26F48.7000205@astro.rug.nl> Message-ID: Thats 2GB. I am guessing your machine is 32bit only? What is your architecture? Brock Palen www.umich.edu/~brockp Center for Advanced Computing brockp at umich.edu (734)936-1985 On Mar 20, 2008, at 10:06 AM, Seyit Hocuk wrote: > 2147483647 From seyit at astro.rug.nl Thu Mar 20 09:51:13 2008 From: seyit at astro.rug.nl (Seyit Hocuk) Date: Thu, 20 Mar 2008 15:51:13 +0100 Subject: [FLASH-USERS] Maximum number of blocks In-Reply-To: References: <47E26F48.7000205@astro.rug.nl> Message-ID: <47E279E1.6070408@astro.rug.nl> Yes that's 2GB and indeed my machine is 32bit only. I use an intel machine core 2 duo E4500 @ 2.20GHz, cache size: 2048 KB, with 3GB internal memory. Is there a way to increase resolution by circumventing this limit? Kind Regards, Seyit Brock Palen wrote: > Thats 2GB. I am guessing your machine is 32bit only? What is your > architecture? > > Brock Palen > www.umich.edu/~brockp > Center for Advanced Computing > brockp at umich.edu > (734)936-1985 > > > > On Mar 20, 2008, at 10:06 AM, Seyit Hocuk wrote: > >> 2147483647 From dubey at flash.uchicago.edu Thu Mar 20 10:12:23 2008 From: dubey at flash.uchicago.edu (Anshu Dubey) Date: Thu, 20 Mar 2008 10:12:23 -0500 (CDT) Subject: [FLASH-USERS] Maximum number of blocks In-Reply-To: <47E279E1.6070408@astro.rug.nl> References: <47E26F48.7000205@astro.rug.nl> <47E279E1.6070408@astro.rug.nl> Message-ID: <64872.128.135.152.184.1206025943.squirrel@flash.uchicago.edu> You could increase the size of individual blocks and decrease the maxblocks. That also decreases the count of number of cells you have, because the ratio of interior to guardcell improves. However, 90000 does seem like an excessive number of blocks. Maxblocks is the maximum number of blocks on a single processor. You could use more processors and thus reduce the maxblocks proportionately. Anshu > Yes that's 2GB and indeed my machine is 32bit only. I use an intel > machine core 2 duo E4500 @ 2.20GHz, cache size: 2048 KB, with 3GB > internal memory. > > Is there a way to increase resolution by circumventing this limit? > > Kind Regards, > Seyit > > > > Brock Palen wrote: >> Thats 2GB. I am guessing your machine is 32bit only? What is your >> architecture? >> >> Brock Palen >> www.umich.edu/~brockp >> Center for Advanced Computing >> brockp at umich.edu >> (734)936-1985 >> >> >> >> On Mar 20, 2008, at 10:06 AM, Seyit Hocuk wrote: >> >>> 2147483647 > From brockp at umich.edu Thu Mar 20 10:17:07 2008 From: brockp at umich.edu (Brock Palen) Date: Thu, 20 Mar 2008 11:17:07 -0400 Subject: [FLASH-USERS] Maximum number of blocks In-Reply-To: <47E279E1.6070408@astro.rug.nl> References: <47E26F48.7000205@astro.rug.nl> <47E279E1.6070408@astro.rug.nl> Message-ID: <47D0A766-9ED1-4B51-AAED-124745CFAA1E@umich.edu> You can't. I won't go into details but under 32bit applications can not use more than 2gb of memory of the 4gb space provided by 32bit addressing. You don't want to anyway which I will try to explain from a system point of view. The flash people will have to help you with flash. Also the E4500 does support EM64T so they CAN run 64 bit if you install 64bit linux (x86_64) and thats what we run here. Now that 2GB is per-process. If you have a dual core system you can run two processes each, each using 2GB. (total system can not have more than 4gb installed without things like PAE or going to x86_64). But because you only have 3GB of memory in each node, you don't want each process using more than 1.5GB of memory. And trust me flash will use it. If you go over this your system will 'swap' reducing performance by orders of magnitude, and will wear out your hard drives. Now MAXBLOCKS is per cpu max number of blocks it can hold at any one time. If your total problem need to have more blocks, you can lower maxblocks, but increase the number of cpus you run on. Letting the problem have more places to put those blocks. If you only have your one system with 3GB of ram, I am am 80% sure you can not do what you want to do with the hardware you have. Sorry. Others please chime in if I am wrong, I do not like my answer. Brock Palen www.umich.edu/~brockp Center for Advanced Computing brockp at umich.edu (734)936-1985 On Mar 20, 2008, at 10:51 AM, Seyit Hocuk wrote: > Yes that's 2GB and indeed my machine is 32bit only. I use an intel > machine core 2 duo E4500 @ 2.20GHz, cache size: 2048 KB, with 3GB > internal memory. > > Is there a way to increase resolution by circumventing this limit? > > Kind Regards, > Seyit > > > > Brock Palen wrote: >> Thats 2GB. I am guessing your machine is 32bit only? What is >> your architecture? >> >> Brock Palen >> www.umich.edu/~brockp >> Center for Advanced Computing >> brockp at umich.edu >> (734)936-1985 >> >> >> >> On Mar 20, 2008, at 10:06 AM, Seyit Hocuk wrote: >> >>> 2147483647 > > > From seyit at astro.rug.nl Thu Mar 20 12:55:14 2008 From: seyit at astro.rug.nl (Seyit Hocuk) Date: Thu, 20 Mar 2008 18:55:14 +0100 Subject: [FLASH-USERS] Maximum number of blocks In-Reply-To: <47D0A766-9ED1-4B51-AAED-124745CFAA1E@umich.edu> References: <47E26F48.7000205@astro.rug.nl> <47E279E1.6070408@astro.rug.nl> <47D0A766-9ED1-4B51-AAED-124745CFAA1E@umich.edu> Message-ID: <47E2A502.8070106@astro.rug.nl> Thanks alot Brock and everyone else who replied. I think you are indeed completely right. I understand now that these maxblocks are the amount per cpu. This was just the thing that I did not know. So I don't have to increase my maxblocks when I increase my total blocks (which is resolution). Just increase the number of cpu's I use. I have tried it on a super computer cluster and it works! Kind Regards, Seyit. Brock Palen wrote: > You can't. I won't go into details but under 32bit applications can > not use more than 2gb of memory of the 4gb space provided by 32bit > addressing. > > You don't want to anyway which I will try to explain from a system > point of view. The flash people will have to help you with flash. > Also the E4500 does support EM64T so they CAN run 64 bit if you > install 64bit linux (x86_64) and thats what we run here. > > Now that 2GB is per-process. If you have a dual core system you can > run two processes each, each using 2GB. (total system can not have > more than 4gb installed without things like PAE or going to x86_64). > But because you only have 3GB of memory in each node, you don't want > each process using more than 1.5GB of memory. And trust me flash will > use it. If you go over this your system will 'swap' reducing > performance by orders of magnitude, and will wear out your hard drives. > > Now MAXBLOCKS is per cpu max number of blocks it can hold at any one > time. If your total problem need to have more blocks, you can lower > maxblocks, but increase the number of cpus you run on. Letting the > problem have more places to put those blocks. > > If you only have your one system with 3GB of ram, I am am 80% sure you > can not do what you want to do with the hardware you have. Sorry. > > Others please chime in if I am wrong, I do not like my answer. > > Brock Palen > www.umich.edu/~brockp > Center for Advanced Computing > brockp at umich.edu > (734)936-1985 > > > > On Mar 20, 2008, at 10:51 AM, Seyit Hocuk wrote: > >> Yes that's 2GB and indeed my machine is 32bit only. I use an intel >> machine core 2 duo E4500 @ 2.20GHz, cache size: 2048 KB, with 3GB >> internal memory. >> >> Is there a way to increase resolution by circumventing this limit? >> >> Kind Regards, >> Seyit >> >> >> >> Brock Palen wrote: >>> Thats 2GB. I am guessing your machine is 32bit only? What is your >>> architecture? >>> >>> Brock Palen >>> www.umich.edu/~brockp >>> Center for Advanced Computing >>> brockp at umich.edu >>> (734)936-1985 >>> >>> >>> >>> On Mar 20, 2008, at 10:06 AM, Seyit Hocuk wrote: >>> >>>> 2147483647 >> >> >> From latife at astro.rug.nl Mon Mar 24 08:30:16 2008 From: latife at astro.rug.nl (M.A. Latife) Date: Mon, 24 Mar 2008 14:30:16 +0100 Subject: [FLASH-USERS] Running FLASH on AMD Machines Message-ID: <47E7ACE8.5070801@astro.rug.nl> Hi, Dear all, hope you will be fine! I am running FLASH 2.5 on Linux cluster but it is slow while running on multi-machine(20 nodes) as compared to even single Machine(intel machine). I am using Opteron Cluster(AMD Machines) to run my simulations(FLASH 2.5). Can any body tell me why it is so slow on multi machine. what is the possible solution kind regards Latife From nhearn at uchicago.edu Mon Mar 24 11:16:10 2008 From: nhearn at uchicago.edu (Nathan Hearn) Date: Mon, 24 Mar 2008 11:16:10 -0500 Subject: [FLASH-USERS] Running FLASH on AMD Machines In-Reply-To: <47E7ACE8.5070801@astro.rug.nl> References: <47E7ACE8.5070801@astro.rug.nl> Message-ID: <2467fdc0803240916j52e62ee7y1254c84a53a7aa76@mail.gmail.com> Hi Latife, Can you tell us a bit more about the system you are using? Some things that would be helpful are: 1. The operating system name and version (or, if running Linux, the distribution version and kernel version); on a Linux or other Unix-like system, running "uname -a" will give you information about the kernel. 2. The name and version of the Fortran compiler you are using (and the compiler flags from Makefile.h); Makefile.h should list the name of the compiler executable (probably something like "mpif90"), and you may be able to run the executable from the command line with an option like "-v" or "--version" to get the version information. 3. The version of the MPI system (e.g., MPICH 1.2.7p1); you may be able to use "mpirun -v" or "mpirun --version" to determine this. 4. The type of network being used by MPI on the cluster (e.g., Ethernet or Myrinet). The system manager should be able to tell you what most of these are if you can't find them. I'm not sure if I can determine the source of the problem, but someone on the users list should be able to use this information to help you. (You could also tell us a bit about the problem you are running, as well as provide the setup command you are using, and maybe the flash.par file.) - Nathan -- Nathan C. Hearn nhearn at uchicago.edu ASC Flash Center Computational Physics Group University of Chicago On Mon, Mar 24, 2008 at 8:30 AM, M.A. Latife wrote: > Hi, > Dear all, > hope you will be fine! > I am running FLASH 2.5 on Linux cluster but it is slow while running > on multi-machine(20 nodes) as compared to even single Machine(intel > machine). > I am using Opteron Cluster(AMD Machines) to run my simulations(FLASH 2.5). > Can any body tell me why it is so slow on multi machine. what is the > possible solution > kind regards > Latife > From latife at astro.rug.nl Mon Mar 24 11:49:49 2008 From: latife at astro.rug.nl (M.A. Latife) Date: Mon, 24 Mar 2008 17:49:49 +0100 Subject: [FLASH-USERS] Running FLASH on AMD Machines In-Reply-To: <2467fdc0803240916j52e62ee7y1254c84a53a7aa76@mail.gmail.com> References: <47E7ACE8.5070801@astro.rug.nl> <2467fdc0803240916j52e62ee7y1254c84a53a7aa76@mail.gmail.com> Message-ID: <47E7DBAD.7040908@astro.rug.nl> Hi Nathan, Thanks for replying so quickly. I will answer each of your questions. These are the specifications for the pc's of the computer cluster. > 1. The operating system name and version (or, if running Linux, the > distribution version and kernel version); on a Linux or other > Unix-like system, running "uname -a" will give you information about > the kernel. > uname -a Linux hpcibm1 2.6.8-11-amd64-k8-smp #1 SMP Sun Oct 2 20:03:22 UTC 2005 x86_64 GNU/Linux > 2. The name and version of the Fortran compiler you are using (and the > compiler flags from Makefile.h); Makefile.h should list the name of > the compiler executable (probably something like "mpif90"), and you > may be able to run the executable from the command line with an option > like "-v" or "--version" to get the version information. > Compiler: pgf90 6.0-5 32-bit target on x86-64 Linux Flags: FFLAGS_OPT = -c -fast -r8 -i4 FFLAGS_DEBUG = -g -c -Mbounds -r8 -i4 FFLAGS_TEST = -c -r8 -i4 -fast -Mprof=lines F90FLAGS = CFLAGS_OPT = -O2 -c CFLAGS_DEBUG = -g -c CFLAGS_TEST = -c > 3. The version of the MPI system (e.g., MPICH 1.2.7p1); you may be > able to use "mpirun -v" or "mpirun --version" to determine this. > I am using mpich2, "mpich2-gcc-1.0.5p4" > 4. The type of network being used by MPI on the cluster (e.g., > Ethernet or Myrinet). > I think just ethernet If anyone knows or have had this sort of problem, please share with me. I think that I am using serial HDF5, maybe this is the problem? kind regards Latife From nhearn at uchicago.edu Mon Mar 24 12:21:36 2008 From: nhearn at uchicago.edu (Nathan Hearn) Date: Mon, 24 Mar 2008 12:21:36 -0500 Subject: [FLASH-USERS] Running FLASH on AMD Machines In-Reply-To: <47E7DBAD.7040908@astro.rug.nl> References: <47E7ACE8.5070801@astro.rug.nl> <2467fdc0803240916j52e62ee7y1254c84a53a7aa76@mail.gmail.com> <47E7DBAD.7040908@astro.rug.nl> Message-ID: <2467fdc0803241021t2e0fbdd6r5ed2bd4e927a0d45@mail.gmail.com> On Mon, Mar 24, 2008 at 11:49 AM, M.A. Latife wrote: > If anyone knows or have had this sort of problem, please share with me. > I think that I am using serial HDF5, maybe this is the problem? Indeed, the use of serial HDF5 could be relevant. From the simulation log file, can you tell if the file reads and/or writes are taking a lot of time? Running with parallel HDF5 would be worth trying. - Nathan From latife at astro.rug.nl Mon Mar 24 12:59:16 2008 From: latife at astro.rug.nl (M.A. Latife) Date: Mon, 24 Mar 2008 18:59:16 +0100 Subject: [FLASH-USERS] Running FLASH on AMD Machines In-Reply-To: <2467fdc0803241021t2e0fbdd6r5ed2bd4e927a0d45@mail.gmail.com> References: <47E7ACE8.5070801@astro.rug.nl> <2467fdc0803240916j52e62ee7y1254c84a53a7aa76@mail.gmail.com> <47E7DBAD.7040908@astro.rug.nl> <2467fdc0803241021t2e0fbdd6r5ed2bd4e927a0d45@mail.gmail.com> Message-ID: <47E7EBF4.10602@astro.rug.nl> Hi Dear Nathan Hearn, I have checked the file(hdf5) write open and close time but it is too little like 2 seconds. cheers latife Nathan Hearn wrote: > On Mon, Mar 24, 2008 at 11:49 AM, M.A. Latife wrote: > >> If anyone knows or have had this sort of problem, please share with me. >> I think that I am using serial HDF5, maybe this is the problem? >> > > Indeed, the use of serial HDF5 could be relevant. From the simulation > log file, can you tell if the file reads and/or writes are taking a > lot of time? > > Running with parallel HDF5 would be worth trying. > > > - Nathan > From hudson at mcs.anl.gov Mon Mar 24 14:16:00 2008 From: hudson at mcs.anl.gov (Randy Hudson) Date: Mon, 24 Mar 2008 14:16:00 -0500 Subject: [FLASH-USERS] VisIt 1.8.x References: <64570.128.135.152.184.1205939262.squirrel@flash.uchicago.edu> Message-ID: <3194280F-C151-4080-A7E0-85372A3EE0CF@mcs.anl.gov> Hi, I maintain the flash visit plugin at the Flash Center (among other things). I've learned that, at least on some flash data, VisIt 1.8.x operates uselessly slowly. E.g., a vector plot I did of a cylindrical subset of the lowest refinement level of a 7.5GB flash file took 14 times as long w/ visit 1.8.0 as w/ visit 1.7.1. I've verified that the latest version of the plugin (that supports file format version 9 of FLASH3) can be built and used w/ visit 1.7.1. If you run into the problem I encountered, I recommend you fall back to 1.7.1 and build the plugin with that. Let me know if you have any problems. Randy Hudson. From amidala at utk.edu Wed Mar 26 13:31:23 2008 From: amidala at utk.edu (espresso) Date: Wed, 26 Mar 2008 14:31:23 -0400 Subject: [FLASH-USERS] Where to compile new programs and modules in FLASH Message-ID: <47F3CFF9@webmail.utk.edu> Hello, I am new to Flash so I apologize for the convoluted nature of this question. I am trying to add a new nuclear reaction network to Flash and I am starting with the aprox13 code as a model. I am currently modifying FLASH2.5/source/source_terms/burn/aprox13_modified/net.F90. I need net.F90 to call a function called read_nuclear_data and use a module called nuclear_data that sits in a program called data.f in the same lowest directory with net.F90. (not the stub net.F90 in burn, but the net.F90 that sits in aprox13_modified under burn) I thought that modifying the Makefile in FLASH2.5/source/source_terms/burn, where net.F90?s parent stub gets instructed to compile would disrupt the modular nature of the program, so I wrote a new Makefile in ?~burn/aprox13_modified? to compile data.f . This works if I only want to do operations within data.f , however when I try to use modules in data.f in the lowest net.F90 I get a compiler error that states, ? Error in opening the compiled module file. Check INCLUDE paths. [NUCLEAR_DATA]?. I?m not sure what ?Check INCLUDE paths? means in this context. The problem may be that data.f needs to compile before net.F90, but I can?t figure out how to do this and preserve modular structure of Flash. Should I move data.f up a directory into Burn and compile it using commands in the Makefile there, where net.F90?s parent stub is compiled ? Is there a way that I can do it through a Makefile in the aprox13_modified directory below Burn using ?INCLUDE paths? or other means? I do have a flash2 manual and I have dug though it to frustration. On page 44 it tells me to ?consult the authors of FLASH? when attempting to add new solvers or physics so I am following that suggestion now. Any help would be appreciated. Thank you, Suzanne Parete-Koon From nhearn at uchicago.edu Wed Mar 26 14:13:33 2008 From: nhearn at uchicago.edu (Nathan Hearn) Date: Wed, 26 Mar 2008 14:13:33 -0500 Subject: [FLASH-USERS] Where to compile new programs and modules in FLASH In-Reply-To: <47F3CFF9@webmail.utk.edu> References: <47F3CFF9@webmail.utk.edu> Message-ID: <2467fdc0803261213y49b5096dq9f48d569efa7aaac@mail.gmail.com> Hi Suzanne, Since it sounds like you are just changing files in (and adding files to) an existing Flash module, you might want to take advantage of the setup routine's automatic file replacement, rather than write a new module. When setup runs, it takes note of the files in your problem directory in FLASH2.5/setups. If any file in the Flash source tree has the same name as a file in the problem directory, the file from the problem directory is used in its place. In other words, if you have a file named net.F90 in the directory FLASH2.5/setups/detonation (which uses the aprox13 module), setup will use this file instead of the file source/source_terms/burn/aprox13/net.F90 when building the object directory for the detonation problem. Files that do not replace a Flash source file can also be added to the problem directory; editing the Makefile in the problem directory will allow you to tell setup that these files should also be compiled. If you wish to use the same modifications for more than one problem, then you may want to create a new module in the Flash source tree. However, using the above method means that you will not have to make any changes to the Flash source, avoiding any of the issues specific to writing new modules; this could also be a good way to prototype the code for a new module. If you found any of this confusing, please let me know. - Nathan -- Nathan C. Hearn nhearn at uchicago.edu ASC Flash Center Computational Physics Group University of Chicago On Wed, Mar 26, 2008 at 1:31 PM, espresso wrote: > Hello, > > I am new to Flash so I apologize for the convoluted nature of this question. I > am trying to add a new > nuclear reaction network to Flash and I am starting with the aprox13 code as a > model. I am currently > modifying FLASH2.5/source/source_terms/burn/aprox13_modified/net.F90. > > I need net.F90 to call a function called read_nuclear_data and use a module > called nuclear_data that > sits in a program called data.f in the same lowest directory with net.F90. > (not the stub net.F90 in burn, > but the net.F90 that sits in aprox13_modified under burn) > I thought that modifying the Makefile in FLASH2.5/source/source_terms/burn, > where net.F90's parent > stub gets instructed to compile would disrupt the modular nature of the > program, so I wrote a new > Makefile in "~burn/aprox13_modified" to compile data.f . This works if I only > want to do operations > within data.f , however when I try to use modules in data.f in the lowest > net.F90 I get a compiler error > that states, " Error in opening the compiled module file. Check INCLUDE > paths. [NUCLEAR_DATA]". I'm > not sure what "Check INCLUDE paths" means in this context. > > The problem may be that data.f needs to compile before net.F90, but I can't > figure out how to do this > and preserve modular structure of Flash. Should I move data.f up a directory > into Burn and compile it > using commands in the Makefile there, where net.F90's parent stub is compiled > ? Is there a way that I > can do it through a Makefile in the aprox13_modified directory below Burn > using "INCLUDE paths" or > other means? > > I do have a flash2 manual and I have dug though it to frustration. On page 44 > it tells me to "consult the > authors of FLASH" when attempting to add new solvers or physics so I am > following that suggestion > now. > > Any help would be appreciated. > > Thank you, > > Suzanne Parete-Koon From lynnreid at flash.uchicago.edu Wed Mar 26 16:31:24 2008 From: lynnreid at flash.uchicago.edu (Lynn Reid) Date: Wed, 26 Mar 2008 16:31:24 -0500 Subject: [FLASH-USERS] Where to compile new programs and modules in FLASH In-Reply-To: <47F3CFF9@webmail.utk.edu> References: <47F3CFF9@webmail.utk.edu> Message-ID: <47EAC0AC.8010900@flash.uchicago.edu> Hi Suzanne: I agree with Nathan that the easiest way to make modifications to a setup is to work within the setups/simulation directory -- this is true within FLASH2.? or FLASH3. I would STRONGLY encourage you to work with FLASH3 if you are going to make more lasting modifications to a nuclear network -- for example, if you want to use your modifications in other simulations. The nuclear burning routines have been extensively reorganized in FLASH3 and hopefully their layout will be much easier to use. The aprox13 module exists in FLASH3 as physics/sourceTerms/Burn/BurnMain/Aprox13. Let us know if you want any assistance with the transition from FLASH2.5 to FLASH3. Lynn espresso wrote: > Hello, > > I am new to Flash so I apologize for the convoluted nature of this question. I > am trying to add a new > nuclear reaction network to Flash and I am starting with the aprox13 code as a > model. I am currently > modifying FLASH2.5/source/source_terms/burn/aprox13_modified/net.F90. > > I need net.F90 to call a function called read_nuclear_data and use a module > called nuclear_data that > sits in a program called data.f in the same lowest directory with net.F90. > (not the stub net.F90 in burn, > but the net.F90 that sits in aprox13_modified under burn) > I thought that modifying the Makefile in FLASH2.5/source/source_terms/burn, > where net.F90?s parent > stub gets instructed to compile would disrupt the modular nature of the > program, so I wrote a new > Makefile in ?~burn/aprox13_modified? to compile data.f . This works if I only > want to do operations > within data.f , however when I try to use modules in data.f in the lowest > net.F90 I get a compiler error > that states, ? Error in opening the compiled module file. Check INCLUDE > paths. [NUCLEAR_DATA]?. I?m > not sure what ?Check INCLUDE paths? means in this context. > > The problem may be that data.f needs to compile before net.F90, but I can?t > figure out how to do this > and preserve modular structure of Flash. Should I move data.f up a directory > into Burn and compile it > using commands in the Makefile there, where net.F90?s parent stub is compiled > ? Is there a way that I > can do it through a Makefile in the aprox13_modified directory below Burn > using ?INCLUDE paths? or > other means? > > I do have a flash2 manual and I have dug though it to frustration. On page 44 > it tells me to ?consult the > authors of FLASH? when attempting to add new solvers or physics so I am > following that suggestion > now. > > Any help would be appreciated. > > Thank you, > > Suzanne Parete-Koon > > -- Dr. Lynn B. Reid, Scientific Programmer ASC Center for Astrophysical Thermonuclear Flashes The University of Chicago 5640 S. Ellis, RI 473; Chicago, IL 60637 USA +1-773-702-0624 phone From bernalcg at astroscu.unam.mx Wed Mar 26 17:41:57 2008 From: bernalcg at astroscu.unam.mx (Giovanny Bernal) Date: Wed, 26 Mar 2008 15:41:57 -0700 Subject: [FLASH-USERS] Cooling In-Reply-To: <47EAC0AC.8010900@flash.uchicago.edu> References: <47F3CFF9@webmail.utk.edu> <47EAC0AC.8010900@flash.uchicago.edu> Message-ID: <2FFE8A4E-E738-4878-897C-3B4B7211BBEF@astroscu.unam.mx> Hi people, The cooling module is already in FLASH3.0?? thanks El 26/03/2008, a las 02:31 PM, Lynn Reid escribi?: > Hi Suzanne: > > I agree with Nathan that the easiest way to make modifications to a > setup is to work within the setups/simulation directory -- this is > true within FLASH2.? or FLASH3. > > I would STRONGLY encourage you to work with FLASH3 if you are going > to make more lasting modifications to a nuclear network -- for > example, if you want to use your modifications in other > simulations. The nuclear burning routines have been extensively > reorganized in FLASH3 and hopefully their layout will be much > easier to use. The aprox13 module exists in FLASH3 as physics/ > sourceTerms/Burn/BurnMain/Aprox13. Let us know if you want any > assistance with the transition from FLASH2.5 to FLASH3. > > Lynn > > espresso wrote: >> Hello, >> >> I am new to Flash so I apologize for the convoluted nature of this >> question. I am trying to add a new nuclear reaction network to >> Flash and I am starting with the aprox13 code as a model. I am >> currently modifying FLASH2.5/source/source_terms/burn/ >> aprox13_modified/net.F90. >> >> I need net.F90 to call a function called read_nuclear_data and use >> a module called nuclear_data that sits in a program called data.f >> in the same lowest directory with net.F90. (not the stub net.F90 >> in burn, but the net.F90 that sits in aprox13_modified under burn) >> I thought that modifying the Makefile in FLASH2.5/source/ >> source_terms/burn, where net.F90?s parent stub gets instructed to >> compile would disrupt the modular nature of the program, so I >> wrote a new Makefile in ?~burn/aprox13_modified? to compile >> data.f . This works if I only want to do operations within >> data.f , however when I try to use modules in data.f in the >> lowest net.F90 I get a compiler error that states, ? Error in >> opening the compiled module file. Check INCLUDE paths. >> [NUCLEAR_DATA]?. I?m not sure what ?Check INCLUDE paths? means in >> this context. >> >> The problem may be that data.f needs to compile before net.F90, >> but I can?t figure out how to do this and preserve modular >> structure of Flash. Should I move data.f up a directory into Burn >> and compile it using commands in the Makefile there, where >> net.F90?s parent stub is compiled ? Is there a way that I can do >> it through a Makefile in the aprox13_modified directory below Burn >> using ?INCLUDE paths? or other means? >> >> I do have a flash2 manual and I have dug though it to frustration. >> On page 44 it tells me to ?consult the authors of FLASH? when >> attempting to add new solvers or physics so I am following that >> suggestion now. >> >> Any help would be appreciated. >> >> Thank you, >> >> Suzanne Parete-Koon >> >> > > > -- > Dr. Lynn B. Reid, Scientific Programmer > ASC Center for Astrophysical Thermonuclear Flashes The University > of Chicago > 5640 S. Ellis, RI 473; Chicago, IL 60637 USA > +1-773-702-0624 phone From dubey at flash.uchicago.edu Thu Mar 27 09:52:11 2008 From: dubey at flash.uchicago.edu (Anshu Dubey) Date: Thu, 27 Mar 2008 09:52:11 -0500 (CDT) Subject: [FLASH-USERS] Cooling In-Reply-To: <2FFE8A4E-E738-4878-897C-3B4B7211BBEF@astroscu.unam.mx> References: <47F3CFF9@webmail.utk.edu> <47EAC0AC.8010900@flash.uchicago.edu> <2FFE8A4E-E738-4878-897C-3B4B7211BBEF@astroscu.unam.mx> Message-ID: <4919.75.3.140.66.1206629531.squirrel@flash.uchicago.edu> Hi Giovanny, Cooling is available as an add-on unit for FLASH3. The reason is that we can only do limited testing on it, and therefore cannot make any strong statement about how well it works because we don't have a good test setup. All we can say is it is FLASH3 compatible and compiles. If you have a good test problem, please let us know. If we can test it, we will include it in the main tarball. Anshu > Hi people, > The cooling module is already in FLASH3.0?? > thanks > > El 26/03/2008, a las 02:31 PM, Lynn Reid escribi?: > >> Hi Suzanne: >> >> I agree with Nathan that the easiest way to make modifications to a >> setup is to work within the setups/simulation directory -- this is >> true within FLASH2.? or FLASH3. >> >> I would STRONGLY encourage you to work with FLASH3 if you are going >> to make more lasting modifications to a nuclear network -- for >> example, if you want to use your modifications in other >> simulations. The nuclear burning routines have been extensively >> reorganized in FLASH3 and hopefully their layout will be much >> easier to use. The aprox13 module exists in FLASH3 as physics/ >> sourceTerms/Burn/BurnMain/Aprox13. Let us know if you want any >> assistance with the transition from FLASH2.5 to FLASH3. >> >> Lynn >> >> espresso wrote: >>> Hello, >>> >>> I am new to Flash so I apologize for the convoluted nature of this >>> question. I am trying to add a new nuclear reaction network to >>> Flash and I am starting with the aprox13 code as a model. I am >>> currently modifying FLASH2.5/source/source_terms/burn/ >>> aprox13_modified/net.F90. >>> >>> I need net.F90 to call a function called read_nuclear_data and use >>> a module called nuclear_data that sits in a program called data.f >>> in the same lowest directory with net.F90. (not the stub net.F90 >>> in burn, but the net.F90 that sits in aprox13_modified under burn) >>> I thought that modifying the Makefile in FLASH2.5/source/ >>> source_terms/burn, where net.F90?s parent stub gets instructed to >>> compile would disrupt the modular nature of the program, so I >>> wrote a new Makefile in ?~burn/aprox13_modified? to compile >>> data.f . This works if I only want to do operations within >>> data.f , however when I try to use modules in data.f in the >>> lowest net.F90 I get a compiler error that states, ? Error in >>> opening the compiled module file. Check INCLUDE paths. >>> [NUCLEAR_DATA]?. I?m not sure what ?Check INCLUDE paths? means in >>> this context. >>> >>> The problem may be that data.f needs to compile before net.F90, >>> but I can?t figure out how to do this and preserve modular >>> structure of Flash. Should I move data.f up a directory into Burn >>> and compile it using commands in the Makefile there, where >>> net.F90?s parent stub is compiled ? Is there a way that I can do >>> it through a Makefile in the aprox13_modified directory below Burn >>> using ?INCLUDE paths? or other means? >>> >>> I do have a flash2 manual and I have dug though it to frustration. >>> On page 44 it tells me to ?consult the authors of FLASH? when >>> attempting to add new solvers or physics so I am following that >>> suggestion now. >>> >>> Any help would be appreciated. >>> >>> Thank you, >>> >>> Suzanne Parete-Koon >>> >>> >> >> >> -- >> Dr. Lynn B. Reid, Scientific Programmer >> ASC Center for Astrophysical Thermonuclear Flashes The University >> of Chicago >> 5640 S. Ellis, RI 473; Chicago, IL 60637 USA >> +1-773-702-0624 phone > From phy1erp at leeds.ac.uk Thu Mar 27 11:59:18 2008 From: phy1erp at leeds.ac.uk (Elliot Parkin) Date: Thu, 27 Mar 2008 16:59:18 +0000 Subject: [FLASH-USERS] Probelms getting Flash3.0 to execute Message-ID: <1206637158.eb4ca6c45ffdc@webmail8.leeds.ac.uk> Hello, I am attempting to run flash3.0 on a Linux machine but when I try and run the code I get this error: problem with execution of flash3 on linast14: [Errno 2] No such file or directory The command I used to execute flash was: mpirun -np 1 flash3 To setup the test case I used: ./setup Sedov -auto +noio Then gmake was used to compile the code which gave SUCCESS but there was these warnings during the make process: Grid_releaseBlkPtr.F90:104: warning: extra tokens at end of #endif directive ... and, In file quicksort_index.F90:93 if (jstack.gt.NSTACK) pause 'NSTACK too small in indexx' 1 Warning: Obsolete: PAUSE statement at (1) In file quicksort_index.F90:204 if (jstack.gt.NSTACK) pause 'NSTACK too small in indexx' 1 Warning: Obsolete: PAUSE statement at (1) The Makefile.h I've constructed looks like this: # Library specific linking # # If a FLASH module has a 'LIBRARY xxx' line in its Config file, we need to # create a macro in this Makefile.h for LIB_xxx, which will be added to the # link line when FLASH is built. This allows us to switch between different # (incompatible) libraries. We also create a _OPT, _DEBUG, and _TEST # library macro to add any performance-minded libraries (like fast math), # depending on how FLASH was setup. #---------------------------------------------------------------------------- LIB_HDF5 = $(HDF5_PATH)/lib -lhdf5 -lz LIB_PNG = -lpng -lz LIB_MPI = LIB_NCMPI = LIB_MPE = #---------------------------------------------------------------------------- # Additional machine-dependent object files # # Add any machine specific files here -- they will be compiled and linked # when FLASH is built. #---------------------------------------------------------------------------- MACHOBJ = #---------------------------------------------------------------------------- # Additional commands #---------------------------------------------------------------------------- MV = mv -f AR = ar -r RM = rm -f CD = cd RL = ranlib ECHO = echo Do you have any idea what may be wrong? Many thanks, Ross Parkin Postgraduate Student School of Physics and Astronomy University of Leeds, UK From seyit at astro.rug.nl Thu Mar 27 12:25:31 2008 From: seyit at astro.rug.nl (Seyit Hocuk) Date: Thu, 27 Mar 2008 18:25:31 +0100 Subject: [FLASH-USERS] Probelms getting Flash3.0 to execute In-Reply-To: <1206637158.eb4ca6c45ffdc@webmail8.leeds.ac.uk> References: <1206637158.eb4ca6c45ffdc@webmail8.leeds.ac.uk> Message-ID: <47EBD88B.3090401@astro.rug.nl> Hi Ross, I am not sure but isn't your command wrong? mpirun -np 1 flash3 First of all shouldn't you do ./flash3. Otherwise, just check your object directory if there even exists a flash3 executable. Greetz, Seyit Elliot Parkin wrote: > Hello, > > I am attempting to run flash3.0 on a Linux machine but when I try and run the > code I get this error: > > problem with execution of flash3 on linast14: [Errno 2] No such file or > directory > > The command I used to execute flash was: > > mpirun -np 1 flash3 > > To setup the test case I used: ./setup Sedov -auto +noio > > Then gmake was used to compile the code which gave SUCCESS but there was these > warnings during the make process: > > Grid_releaseBlkPtr.F90:104: warning: extra tokens at end of #endif directive > > ... and, > > In file quicksort_index.F90:93 > > if (jstack.gt.NSTACK) pause 'NSTACK too small in indexx' > 1 > Warning: Obsolete: PAUSE statement at (1) > In file quicksort_index.F90:204 > > if (jstack.gt.NSTACK) pause 'NSTACK too small in indexx' > 1 > Warning: Obsolete: PAUSE statement at (1) > > > The Makefile.h I've constructed looks like this: > > # Library specific linking > # > # If a FLASH module has a 'LIBRARY xxx' line in its Config file, we need to > # create a macro in this Makefile.h for LIB_xxx, which will be added to the > # link line when FLASH is built. This allows us to switch between different > # (incompatible) libraries. We also create a _OPT, _DEBUG, and _TEST > # library macro to add any performance-minded libraries (like fast math), > # depending on how FLASH was setup. > #---------------------------------------------------------------------------- > > LIB_HDF5 = $(HDF5_PATH)/lib -lhdf5 -lz > LIB_PNG = -lpng -lz > > LIB_MPI = > LIB_NCMPI = > LIB_MPE = > > > #---------------------------------------------------------------------------- > # Additional machine-dependent object files > # > # Add any machine specific files here -- they will be compiled and linked > # when FLASH is built. > #---------------------------------------------------------------------------- > > MACHOBJ = > > #---------------------------------------------------------------------------- > # Additional commands > #---------------------------------------------------------------------------- > > MV = mv -f > AR = ar -r > RM = rm -f > CD = cd > RL = ranlib > ECHO = echo > > > Do you have any idea what may be wrong? > > Many thanks, > > Ross Parkin > Postgraduate Student > School of Physics and Astronomy > University of Leeds, UK > From dubey at flash.uchicago.edu Thu Mar 27 12:27:35 2008 From: dubey at flash.uchicago.edu (Anshu Dubey) Date: Thu, 27 Mar 2008 12:27:35 -0500 (CDT) Subject: [FLASH-USERS] Probelms getting Flash3.0 to execute In-Reply-To: <1206637158.eb4ca6c45ffdc@webmail8.leeds.ac.uk> References: <1206637158.eb4ca6c45ffdc@webmail8.leeds.ac.uk> Message-ID: <64960.128.135.152.184.1206638855.squirrel@flash.uchicago.edu> It sounds like your executable is in a different directory from where you are trying to execute. The FLASH build process creates a directory called "object" as a default, and the executable is placed there. Try changing the directory to object, and give the run command again. If you still have problem, please let us know. Anshu > Hello, > > I am attempting to run flash3.0 on a Linux machine but when I try and run > the > code I get this error: > > problem with execution of flash3 on linast14: [Errno 2] No such file or > directory > > The command I used to execute flash was: > > mpirun -np 1 flash3 > > To setup the test case I used: ./setup Sedov -auto +noio > > Then gmake was used to compile the code which gave SUCCESS but there was > these > warnings during the make process: > > Grid_releaseBlkPtr.F90:104: warning: extra tokens at end of #endif > directive > > ... and, > > In file quicksort_index.F90:93 > > if (jstack.gt.NSTACK) pause 'NSTACK too small in indexx' > 1 > Warning: Obsolete: PAUSE statement at (1) > In file quicksort_index.F90:204 > > if (jstack.gt.NSTACK) pause 'NSTACK too small in indexx' > 1 > Warning: Obsolete: PAUSE statement at (1) > > > The Makefile.h I've constructed looks like this: > > # Library specific linking > # > # If a FLASH module has a 'LIBRARY xxx' line in its Config file, we need > to > # create a macro in this Makefile.h for LIB_xxx, which will be added to > the > # link line when FLASH is built. This allows us to switch between > different > # (incompatible) libraries. We also create a _OPT, _DEBUG, and _TEST > # library macro to add any performance-minded libraries (like fast math), > # depending on how FLASH was setup. > #---------------------------------------------------------------------------- > > LIB_HDF5 = $(HDF5_PATH)/lib -lhdf5 -lz > LIB_PNG = -lpng -lz > > LIB_MPI = > LIB_NCMPI = > LIB_MPE = > > > #---------------------------------------------------------------------------- > # Additional machine-dependent object files > # > # Add any machine specific files here -- they will be compiled and linked > # when FLASH is built. > #---------------------------------------------------------------------------- > > MACHOBJ = > > #---------------------------------------------------------------------------- > # Additional commands > #---------------------------------------------------------------------------- > > MV = mv -f > AR = ar -r > RM = rm -f > CD = cd > RL = ranlib > ECHO = echo > > > Do you have any idea what may be wrong? > > Many thanks, > > Ross Parkin > Postgraduate Student > School of Physics and Astronomy > University of Leeds, UK > From nhearn at uchicago.edu Thu Mar 27 12:33:55 2008 From: nhearn at uchicago.edu (Nathan Hearn) Date: Thu, 27 Mar 2008 12:33:55 -0500 Subject: [FLASH-USERS] Probelms getting Flash3.0 to execute In-Reply-To: <47EBD88B.3090401@astro.rug.nl> References: <1206637158.eb4ca6c45ffdc@webmail8.leeds.ac.uk> <47EBD88B.3090401@astro.rug.nl> Message-ID: <2467fdc0803271033w53f5ec2bk34c2c721c0cc3299@mail.gmail.com> On Thu, Mar 27, 2008 at 12:25 PM, Seyit Hocuk wrote: > First of all shouldn't you do ./flash3. Actually, while simply running an MPI executable without mpirun does work with some MPI implementations (e.g., OpenMPI and LAM-MPI), others will require mpirun even if only one process is being requested. But in any case, locating the flash3 binary is a good suggestion. - Nathan -- Nathan C. Hearn nhearn at uchicago.edu ASC Flash Center Computational Physics Group University of Chicago From seyit at astro.rug.nl Thu Mar 27 12:37:45 2008 From: seyit at astro.rug.nl (Seyit Hocuk) Date: Thu, 27 Mar 2008 18:37:45 +0100 Subject: [FLASH-USERS] Probelms getting Flash3.0 to execute In-Reply-To: <2467fdc0803271033w53f5ec2bk34c2c721c0cc3299@mail.gmail.com> References: <1206637158.eb4ca6c45ffdc@webmail8.leeds.ac.uk> <47EBD88B.3090401@astro.rug.nl> <2467fdc0803271033w53f5ec2bk34c2c721c0cc3299@mail.gmail.com> Message-ID: <47EBDB69.1000601@astro.rug.nl> Hi Nathan, I actually meant: mpirun -np x ./flash2 instead of mpirun -np x flash2. I am not sure, but I always do ./ in front of executables. Nathan Hearn wrote: > On Thu, Mar 27, 2008 at 12:25 PM, Seyit Hocuk wrote: > >> First of all shouldn't you do ./flash3. >> > > Actually, while simply running an MPI executable without mpirun does > work with some MPI implementations (e.g., OpenMPI and LAM-MPI), others > will require mpirun even if only one process is being requested. > > But in any case, locating the flash3 binary is a good suggestion. > > > - Nathan > > > From phy1erp at leeds.ac.uk Sun Mar 30 06:30:30 2008 From: phy1erp at leeds.ac.uk (Elliot Parkin) Date: Sun, 30 Mar 2008 12:30:30 +0100 Subject: [FLASH-USERS] Problems getting Flash3.0 to execute Message-ID: <1206876630.12005e2535d33@webmail7.leeds.ac.uk> Hello, Thanks for all the suggestions everyone. I forgot to mention that after using the setup command I changed to the object directory. The flash3 binary is definitely there and using the execute command, mpirun -np 1 ./flash3 now gives the error, [cli_0]: aborting job: application called MPI_Abort(MPI_COMM_WORLD, 1140850688) - process 0 RuntimeParameters_read: ignoring unknown parameter "basenm"... RuntimeParameters_read: ignoring unknown parameter "checkpointFileIntervalTime"... rank 0 in job 2 linast14_40610 caused collective abort of all ranks exit status of rank 0: return code 0 I suspect this has something to do with my mpi setup. I've added a line into my Makefile.h for the libmpich.a and libfmpich.a which reads: LIB_MPI = -L$(MPI_PATH)/lib -lmpich -lfmpich -lz but I'm not sure if I've done this right or not. Thanks for help, Ross Parkin Postgraduate Student School of Physics and Astronomy University of Leeds, UK Quoting Elliot Parkin on Thu 27 Mar 2008 16:59:18 GMT: > Hello, > > I am attempting to run flash3.0 on a Linux machine but when I try and run the > code I get this error: > > problem with execution of flash3 on linast14: [Errno 2] No such file or > directory > > The command I used to execute flash was: > > mpirun -np 1 flash3 > > To setup the test case I used: ./setup Sedov -auto +noio > > Then gmake was used to compile the code which gave SUCCESS but there was > these > warnings during the make process: > > Grid_releaseBlkPtr.F90:104: warning: extra tokens at end of #endif directive > > ... and, > > In file quicksort_index.F90:93 > > if (jstack.gt.NSTACK) pause 'NSTACK too small in indexx' > 1 > Warning: Obsolete: PAUSE statement at (1) > In file quicksort_index.F90:204 > > if (jstack.gt.NSTACK) pause 'NSTACK too small in indexx' > 1 > Warning: Obsolete: PAUSE statement at (1) > > > The Makefile.h I've constructed looks like this: > > # Library specific linking > # > # If a FLASH module has a 'LIBRARY xxx' line in its Config file, we need to > # create a macro in this Makefile.h for LIB_xxx, which will be added to the > # link line when FLASH is built. This allows us to switch between different > # (incompatible) libraries. We also create a _OPT, _DEBUG, and _TEST > # library macro to add any performance-minded libraries (like fast math), > # depending on how FLASH was setup. > #---------------------------------------------------------------------------- > > LIB_HDF5 = $(HDF5_PATH)/lib -lhdf5 -lz > LIB_PNG = -lpng -lz > > LIB_MPI = > LIB_NCMPI = > LIB_MPE = > > > #---------------------------------------------------------------------------- > # Additional machine-dependent object files > # > # Add any machine specific files here -- they will be compiled and linked > # when FLASH is built. > #---------------------------------------------------------------------------- > > MACHOBJ = > > #---------------------------------------------------------------------------- > # Additional commands > #---------------------------------------------------------------------------- > > MV = mv -f > AR = ar -r > RM = rm -f > CD = cd > RL = ranlib > ECHO = echo > > > Do you have any idea what may be wrong? > > Many thanks, > > Ross Parkin > Postgraduate Student > School of Physics and Astronomy > University of Leeds, UK > From flash-users at flash.uchicago.edu Sun Mar 30 17:11:27 2008 From: flash-users at flash.uchicago.edu (flash-users at flash.uchicago.edu) Date: Sun, 30 Mar 2008 17:11:27 -0500 (CDT) Subject: [FLASH-USERS] [SPAM] Message 1356586 Hurry! Final Days to get 73% OFF! Message-ID: <20080331131130.7140.qmail@JoseAntonio> An HTML attachment was scrubbed... URL: http://flash.uchicago.edu/pipermail/flash-users/attachments/20080330/3ea4edfc/attachment.html From josef.stoeckl at uibk.ac.at Mon Mar 31 04:10:05 2008 From: josef.stoeckl at uibk.ac.at (=?ISO-8859-1?Q?Josef_St=F6ckl?=) Date: Mon, 31 Mar 2008 11:10:05 +0200 Subject: [FLASH-USERS] StaggeredMesh MHD Solver for 3-dimensional Problems Message-ID: <47F0AA6D.40905@uibk.ac.at> Hello, I am just setting up a 3D problem with Flash which requires MHD and wanted to use the new StaggeredMesh solver. This solver however doesn't support 3 dimensional problems yet, which it illustrates with the following abort message: [MHD_init]: This version of the StaggeredMesh MHD code does not support 3D setups. Please use the add-on version. Now, I was wondering, where I might get this "add-on version" of the code, since the supplementary units package doesn't contain any Hydro units. Many thanks in advance, Josef Stoeckl Institute for Astro- and Particle Physics University of Innsbruck, Austria -------------- next part -------------- A non-text attachment was scrubbed... Name: josef_stoeckl.vcf Type: text/x-vcard Size: 364 bytes Desc: not available Url : http://flash.uchicago.edu/pipermail/flash-users/attachments/20080331/349a8fe4/attachment-0001.vcf From dongwook at flash.uchicago.edu Mon Mar 31 11:53:18 2008 From: dongwook at flash.uchicago.edu (Dongwook Lee) Date: Mon, 31 Mar 2008 11:53:18 -0500 (CDT) Subject: [FLASH-USERS] StaggeredMesh MHD Solver for 3-dimensional Problems In-Reply-To: <47F0AA6D.40905@uibk.ac.at> References: <47F0AA6D.40905@uibk.ac.at> Message-ID: Hi Josef, I am the author of the Staggered mesh MHD solver. The full 3D version is supported in the "research" mode in this FLASH3 release in order to provide the 3D StaggeredMesh solver with more collaborations. Users can directly send me an email to get the full 3D version. I will send you a tar ball in a seperate email and you can simply replace the current directory with new ones. Best, Dongwook ---------------------------------------------------- Dongwook Lee, Ph.D., Postdoctoral Scholar ASC Center for Astrophysical Thermonuclear Flashes University of Chicago 5640 S. Ellis, RI 484 (773) 834-6830 On Mon, 31 Mar 2008, Josef St?ckl wrote: > Hello, > > I am just setting up a 3D problem with Flash which requires MHD and wanted to > use the new StaggeredMesh solver. This solver however doesn't support 3 > dimensional problems yet, which it illustrates with the following abort > message: > > [MHD_init]: This version of the StaggeredMesh MHD code does not support 3D > setups. Please use the add-on version. > > Now, I was wondering, where I might get this "add-on version" of the code, > since the supplementary units package doesn't contain any Hydro units. > > Many thanks in advance, > > Josef Stoeckl > Institute for Astro- and Particle Physics > University of Innsbruck, Austria >