Compiling the Einstein Toolkit - Revision history

Rhaas: wipe content and point to ET tutorial to avoid duplication

2020-07-24T15:11:23Z

wipe content and point to ET tutorial to avoid duplication

Rhaas: update documentation to no longer mention the various Linux variants and to refer to the tutorial for the list of packages

2018-08-31T14:44:42Z

update documentation to no longer mention the various Linux variants and to refer to the tutorial for the list of packages

128.84.127.199: /* Supported laptop/workstation operating systems */

2016-06-13T12:43:43Z

‎Supported laptop/workstation operating systems

Hinder: /* Configuring SimFactory for your machine */

2016-06-13T07:43:55Z

‎Configuring SimFactory for your machine

Hinder at 07:30, 13 June 2016

2016-06-13T07:30:04Z

Hinder at 07:24, 13 June 2016

2016-06-13T07:24:24Z

Show changes

Hinder: /* Building the Einstein Toolkit */

2016-06-13T07:18:35Z

‎Building the Einstein Toolkit

Hinder: /* Configuring SimFactory for your machine */

2016-06-13T07:18:17Z

‎Configuring SimFactory for your machine

Hinder: /* Supported clusters */

2016-06-13T07:16:48Z

‎Supported clusters

Hinder: /* Unsupported machines */

2016-06-13T07:14:59Z

‎Unsupported machines

←Older revision		Revision as of 15:11, 24 July 2020
Line 1:		Line 1:
−	~~The recommended way to compile the Einstein Toolkit is to use the [http://simfactory.org Simulation Factory] ("SimFactory").~~	+	Please see https://github.com/nds-org/jupyter-et/blob/master/CactusTutorial.ipynb for build instructions on Linux, OSX and Windows.
−
−	~~==Configuring SimFactory for your machine==~~
−
−	The ET depends on various libraries, and needs to interact with machine-specific queueing systems and MPI implementations. As such, it needs to be configured for a given machine. For this, it uses SimFactory. Generally, configuring SimFactory means providing an optionlist, for specifying library locations and build options, a submit script for using the batch queueing system, and a runscript, for specifying how Cactus should be run, e.g. which mpirun command to use. For a cluster, a "machine definition file" is also required.
−
−	~~===Supported clusters===~~
−
−	~~SimFactory comes with support for many of the large supercomputers; run "simfactory/bin/sim list-machines" from your Cactus directory to~~ see ~~the currently-supported list of machines. If you are compiling on one of these, it should be sufficient to run~~
−
−	~~simfactory/bin/sim setup~~
−
−	~~from your Cactus directory and answer the questions, and SimFactory will automatically detect the correct option list, submit script and run script.~~
−
−	~~===Supported laptop/workstation operating systems===~~
−
−	SimFactory also contains general support for specific operating systems commonly used on workstations or laptops, including Mac OS, Ubuntu, Cent OS and Scientific Linux. To configure SimFactory for one of these, you need to find the suitable files in simfactory/mdb/optionlists and simfactory/mdb/runscripts and specify their names on the sim setup command line. The following table lists the files that you need for each supported operating system:
−
−	~~{\| class="wikitable" style="border: 1pt solid black;"~~
−	~~\|+Configuration files~~
−	\|-
−	~~!Operating system~~
−	~~!Option list~~
−	~~!Run script~~
−	~~!Notes~~
−	\|-
−	~~\|Linux (inc. the Linux subsystem of Windows 10)~~
−	~~\|generic.cfg~~
−	~~\|generic.run~~
−	\|-
−	~~\|OS X (with MacPorts)~~
−	~~\|osx-macports.cfg~~
−	~~\|osx-macports.run~~
−	\|-
−	~~\|OS X (with HomeBrew)~~
−	~~\|osx-homebrew.cfg~~
−	~~\|generic.run~~
−	\|-
−	\|}
−
−	~~You do not need to specify a submit script, as there is no queueing system on a workstation or laptop. For example, for Debian, after installing the packages listed in [~~https://~~nbviewer.jupyter~~.~~org/github~~/nds-org/jupyter-et/blob/master/CactusTutorial.ipynb ~~the offline tutorial],~~
−
−	~~simfactory/bin/sim setup --optionlist=generic.cfg --runscript generic.run~~
−
−	~~will configure SimFactory~~ for ~~your machine.~~
−
−	~~===Unsupported machines===~~
−
−	~~If there is no existing optionlist or runscript (for a cluster, also a submit script) for your machine~~, you will need to write them yourself. It is recommended to start from one of the already-written files for another machine which is as close as possible to the machine you are trying to use. See [[Configuring a new machine]] for help with this. First, write the option list and ~~run script, and run~~
−
−	~~simfactory/bin/sim setup --optionlist=<optionlist> --runscript <runscript>~~
−
−	with the new configuration files. This will create a basic machine definition file in simfactory/mdb/machines named after your hostname. For a cluster, it is unlikely that this basic file will be enough, as SimFactory needs a lot of information about the details of the cluster.
−
−	~~==Building the Einstein Toolkit==~~
−
−	~~Assuming that SimFactory has been successfully set up on your machine, you should be able to build the Einstein Toolkit with~~
−
−	~~simfactory/bin/sim build --thornlist path/to/thornlist~~.th

@@ Line 25: / Line 25: @@
 !Notes
 |-
-|Debian
+|Linux (inc. the Linux subsystem of Windows 10)
-|debian.cfg
+|generic.cfg
-|debian.sh
+|generic.run
 |-
 |OS X (with MacPorts)
@@ Line 44: / Line 35: @@
 |OS X (with HomeBrew)
 |osx-homebrew.cfg
-|generic-mpi.run
+|generic.run
 |-
 |}
-You do not need to specify a submit script, as there is no queueing system on a workstation or laptop.  Often the optionlist will contain a comment near the top explaining which system packages need to be installed to use the optionlist.  For example, for Debian, after installing the packages listed in the comment at the top of debian.cfg,
+You do not need to specify a submit script, as there is no queueing system on a workstation or laptop.  For example, for Debian, after installing the packages listed in [https://nbviewer.jupyter.org/github/nds-org/jupyter-et/blob/master/CactusTutorial.ipynb the offline tutorial],
-     simfactory/bin/sim setup --optionlist=debian.cfg --runscript debian.sh
+     simfactory/bin/sim setup --optionlist=generic.cfg --runscript generic.run
-will configure SimFactory for your machine.  For some systems, the runscript can be the general "generic-mpi.sh", or one from a similar OS, for example for Ubuntu you can use debian.sh as a run script.
+will configure SimFactory for your machine.
 ===Unsupported machines===

@@ Line 45: / Line 45: @@
 |osx-homebrew.cfg
 |generic-mpi.run
-|Run 'export CPATH=/usr/local/include LIBRARY_PATH=/usr/local/lib" before building or running.
+|-
 |}

@@ Line 3: / Line 3: @@
 ==Configuring SimFactory for your machine==
-The ET depends on various libraries, and needs to interact with machine-specific queueing systems and MPI implementations.  As such, it needs to be configured for a given machine.  For this, it uses SimFactory.  Generally, configuring SimFactory means providing an optionlist, for specifying library locations and build options, a submit script for using the batch queueing system, and a runscript, for specifying how Cactus should be run, e.g. which mpirun command to use.
+The ET depends on various libraries, and needs to interact with machine-specific queueing systems and MPI implementations.  As such, it needs to be configured for a given machine.  For this, it uses SimFactory.  Generally, configuring SimFactory means providing an optionlist, for specifying library locations and build options, a submit script for using the batch queueing system, and a runscript, for specifying how Cactus should be run, e.g. which mpirun command to use.  For a cluster, a "machine definition file" is also required.
 ===Supported clusters===

@@ Line 56: / Line 56: @@
 ===Unsupported machines===
-If there is no existing optionlist or runscript (for a cluster, also a submit script) for your machine, you will need to write them yourself.  It is recommended to start from one of the already-written files for another machine which is as close as possible to the machine you are trying to use.  See [[Configuring a new machine]] for help with this. For a workstation or laptop, you only need to provide the option list and run script, and then run
+If there is no existing optionlist or runscript (for a cluster, also a submit script) for your machine, you will need to write them yourself.  It is recommended to start from one of the already-written files for another machine which is as close as possible to the machine you are trying to use.  See [[Configuring a new machine]] for help with this. First, write the option list and run script, and run
      simfactory/bin/sim setup --optionlist=<optionlist> --runscript <runscript>
-with the new configuration files.  For a cluster, it is unlikely that this will be enough, as SimFactory needs a lot of information about the details of the cluster, provided in a "machine definition file" in simfactory/mdb/machines.  Copy an example file, and adapt it to your machine.  Getting this right is nontrivial.  In that case, the machine definition file contains the name of the option list and run script to us, so once the machine definition is complete, you only need to run
+with the new configuration files.  This will create a basic machine definition file in simfactory/mdb/machines named after your hostname.  For a cluster, it is unlikely that this basic file will be enough, as SimFactory needs a lot of information about the details of the cluster.
 ==Building the Einstein Toolkit==

@@ Line 66: / Line 66: @@
 Assuming that SimFactory has been successfully set up on your machine, you should be able to build the Einstein Toolkit with
-     sim build --thornlist path/to/thornlist.th
+     simfactory/bin/sim build --thornlist path/to/thornlist.th
 ==Configuring a new machine==

@@ Line 7: / Line 7: @@
 ===Supported clusters===
-SimFactory comes with support for many of the large supercomputers; run "sim list-machines" from your Cactus directory to see the currently-supported list of machines.  If you are compiling on one of these, it should be sufficient to run
+SimFactory comes with support for many of the large supercomputers; run "simfactory/bin/sim list-machines" from your Cactus directory to see the currently-supported list of machines.  If you are compiling on one of these, it should be sufficient to run
-     sim setup
+     simfactory/bin/sim setup
 from your Cactus directory and answer the questions, and SimFactory will automatically detect the correct option list, submit script and run script.
@@ Line 50: / Line 50: @@
 You do not need to specify a submit script, as there is no queueing system on a workstation or laptop.  Often the optionlist will contain a comment near the top explaining which system packages need to be installed to use the optionlist.  For example, for Debian, after installing the packages listed in the comment at the top of debian.cfg,
-     sim setup --optionlist=debian.cfg --runscript debian.sh
+     simfactory/bin/sim setup --optionlist=debian.cfg --runscript debian.sh
 will configure SimFactory for your machine.  For some systems, the runscript can be the general "generic-mpi.sh", or one from a similar OS, for example for Ubuntu you can use debian.sh as a run script.
@@ Line 58: / Line 58: @@
 If there is no existing optionlist or runscript (for a cluster, also a submit script) for your machine, you will need to write them yourself.  It is recommended to start from one of the already-written files for another machine which is as close as possible to the machine you are trying to use.  See the sections below for help with this.  Once the files are written, run
-     sim setup --optionlist=<optionlist> --runscript <runscript> --submitscript <submitscript>
+     simfactory/bin/sim setup --optionlist=<optionlist> --runscript <runscript> --submitscript <submitscript>
 with the new configuration files.  For a cluster, it is unlikely that this will be enough, as SimFactory needs a lot of information about the details of the cluster, provided in a "machine definition file" in simfactory/mdb/machines.  Copy an example file, and adapt it to your machine.  Getting this right is nontrivial.