Difference between revisions of "Cluster:Gaussian"

From Earlham CS Department
Jump to navigation Jump to search
m (Gaussian Installation)
m (Replaced content with "Go here")
(Tag: Replaced)
Line 1: Line 1:
Web interface: [[WebMO]]
[[Sysadmin:WebMO|Go here]]
Users for WebMo:
* Create an LDAP user.
* [http://w0.cluster.earlham.edu/~alice/webmo-cgi-bin/login.cgi Log in] using that user's credentials
* Give the group password. (This is a one-time requirement. Future logins will not ask for it.)
* Use Gaussian.
== Gaussian Installation ==
The current path to Gaussian: /mounts/whedon/software/Gaussian
The ISO images for Gaussian installer live in /cluster/gaussian. For whedon, I used the ISO image intended for CPUs that support SSE4.2 instructions.
First, mount this image in some directory using mount -o command. After that, follow these instructions: https://gaussian.com/g16/g16bin_install.pdf.
You can connect Gaussian to WebMO through WebMO's web interface (log in as admin and go to Interface Manager). The current version of Gaussian that we are using is g16, but our WebMO installation is so old that it only supports Gaussian up to g09. In order to overcome this problem, we have two symlinks /mounts/whedon/software/Gaussian/g09 and /mounts/whedon/software/Gaussian/g16/g09, pointing to the directories /mounts/whedon/software/Gaussian/g16 and /mounts/whedon/software/Gaussian/g16/g16 respectively. After creating these symlinks, use the 'suggest' button in WebMO's Gaussian Manager that provides the paths for different Gaussian directories for you.
To see a particular job's files, go to /mounts/whedon/webmo-users/$user/$job_number/ and examine the relevant files.
== An old WebMO configuration example ==
[[Cluster:_New_BobSCEd_Install_Log|Notes below taken from here.]]
* yum installed httpd
* Installed on bs0 with the following params:
Path to perl:        /usr/bin/perl
Webserver name:      bs0-new.cluster.earlham.edu
HTML directory:      /var/www/webmo
HTML URL:            /webmo
CGI script directory: /var/www/cgi-bin
CGI script URL:      /cgi-bin
User files directory: /mounts/bobsced/WebMO</pre>
* Get this error when authing with LDAP: <code>Can't locate Authen/Simple/LDAP.pm</code>
* yum installed perl-LDAP.noarch, didn't work, so used CPAN to install Authen::Simple::LDAP
* edited /var/www/cgi-bin/interfaces/authen.conf for our LDAP settings
* Before externally authenticated users can use it, you have to go in as administrator and check the box to allow them in the Webmo group (or whatever other group)
* Gamess:
** yum install compat-gcc-34-g77.x86_64 and gfortran
** Followed directions from [http://www.webmo.net/support/gamess_linux.html Webmo site]
* Added the following line to httpd.conf:
:<code>SuexecUserGroup bob users </code>
* Gaussian 09 not supported, though it's installed in /mounts/bobsced/usr/local/g09
* Installed g03, except get errors:
<pre>Erroneous write during file extend. write 160 instead of 4096
Probably out of disk space.
Write error in NtrExt1: No such file or directory
<pre>Write error in NtrExt1: Bad address</pre>
** To fix this, do <code>echo 0 > /proc/sys/kernel/randomize_va_space</code>
** <font color="green">This needs to be set to happen all the time on boot</font>
== Older notes, c. 2005 ==
* [[Cluster:Gaussian environment|Gaussian environment]]
* [[Cluster:Running Gaussian in parallel|Running Gaussian in parallel]]
* [[Cluster:Gaussian PBS script|Sample Gaussian PBS script]]
=== Gaussian environment ===
I use this in my .bashrc:
        export GAUSS_EXEDIR=/cluster/bazaar/software/g03
        export GAUSS_LFLAGS='-nodelist "b0 b1" -opt "Tsnet.Node.lindarsharg: ssh " -mp 2'
        export GAUSS_SCRDIR=/tmp
        export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/cluster/bazaar/software/g03
        export g03root=/cluster/bazaar/software
        export PATH=$PATH:/cluster/bazaar/software/g03
Seems to work, although you might want to tweak the node list.
=== Running Gaussian in parallel ===
Understand the distinction between <tt>nprocs</tt> and <tt>nprocl</tt>. <tt>nprocs</tt> is for a shared-memory (single-image) machine, whereas <tt>nprocl</tt> is for a distributed-memory (cluster) system. You want to use <tt>nprocl</tt> on all our computing systems. For example, if you want to run on eight processes (four nodes), add a line like this
to your input file. You should then specify <tt>qsub -l nodes=4:ppn=2 ''script''</tt> on the command line to submit your PBS script with a requirement of four dual-processor machines.
Once you have the number of CPUs specified in the input file, make sure that your nodelist in GAUSS_LFLAGS is set properly. See [[Cluster:Gaussian_environment|Gaussian environment]] for details.
If you have a bunch of input files you want to modify, use the add_nproc.pl script in /cluster/generic/bin. First cd to the directory with your input files, and then do:
perl /cluster/generic/bin/add_nproc.pl ''n''
where ''n'' is the number proecssors you want to run on.
After that, fire it up with <tt>$g03root/bsd/g03l</tt>.
=== Sample Gaussian PBS script ===
This is a sample PBS script for Gaussian:
#PBS -N gaussian_test061.com_nodes=2:ppn=2
#PBS -o /cluster/home/skylar/bazaar/gaussian_nodes2/test061.com.out
#PBS -e /cluster/home/skylar/bazaar/gaussian_nodes2/test061.com.err
#PBS -q batch
#PBS -m abe
#PBS -l nodes=2:ppn=2
/cluster/bazaar/software/g03/bsd/g03l < /cluster/bazaar/software/g03/tests/com_smp4/test061.com > /cluster/home/skylar/bazaar/gaussian_nodes2/test061.com.log
exit $!

Latest revision as of 12:40, 22 August 2021