View source

The Darwin compute cluster consists of 128 dual-cpu dual-core nodes. Each cpu is a 3.0GHz Intel Woodcrest processor. Each node has 8GB of memory and a 1333MHz front-side-bus. The compute nodes are connected with a Myrinet-10G network. The compute cluster runs an installation of the [http://microbe.csbi.mit.edu/wordpress/ ROCKS] cluster software. 

== Hardware ==

The dual-core, dual-CPU 3.0Ghz, 8GB compute cluster nodes are IBM x3550 xSeries computers. These are 1U, rack-mounted systems that make a lot of noise and produce a lot of heat. Some details on the IBM system X machines used for the compute cluster can be found [[media:system_x_x3550.pdf | here]]. The front-end, login node for the system is an IBM xSeries x3650 machine. This is the system you access when you first log in to the cluster. It is a 2U equivalent of the compute cluster machines. It is a dual-cpu, dual-core 3.0GHz Woodcrest system with 10GB of main memory. Some details on this type of system can be found [[media:system_x_x3650.pdf | here]]. 

=== Compute hardware organization ===

The front-end, login node is named ''beagle'' (after the ship H.M.S Darwin that Charles Darwin travelled on during his explorations). The compute nodes are named according to the number of the rack they are housed in and according to their rank within that rack. The rack numbers range from 1-4, the compute node
numbers within each rack range from 0-31. This gives compute node names such as
 compute-1-0
for the 0 rank compute node in rack 1. The compute nodes also have shortened names of the form c''N''-''M'', where ''N'' is the rack number and ''M'' the node rank. Each node also has a name of the form compute-''N''-''M''-myri0. This form of name can be used to access a node through its Myrinet switch network interface.

=== High speed network ===

The 128 compute nodes and the [[Storage hardware/software overview | storage infrastructure]] are all linked
to a Myrinet 10G network switch.

== Available Software ==

In addition to the basic [http://microbe.csbi.mit.edu/wordpress/ ROCKS] distribution a number of software packages are available. The full set changes over time, as software package installations are requested and/or upgraded versions become available. To manage this and make software accessible to system users we employ the [http://modules.sourceforge.net/ modules] tool. Using this tool you can see what software is available and configure your cluster account to use specific tools. More details on modules can be found [[software_help:modules | here]]. At present the list of software available using modules is short, but it is growing! The current list can be found by using the command

 [charles@beagle ~]$ module avail
 -------------------------------- /home/admin/modulefiles ---------------------------------
 matlab/7.4.0.287(2007a:default) netcdf/3.6.2

Basic examples and howto information for key pieces of software can be found in the [[software_help | software help]] pages.

== MySQL for user DBs ==

A separate instance of the MySQL server is running on beagle to provide access for users needing to create a DB for their research.  This DB server is installed at a different location and is available on separate sockets and ports from the system server so as to not interfere with the DB server for the rocks cluster.  To use this DB instance load in the mysql module which will define all the necessary system envs for you:

 beagle ~]$ module load mysql

This defines the following envs: '''MYSQL_HOME, MYSQL_HOST, MYSQL_TCP_PORT, MYSQL_UNIX_PORT''' and updates '''PATH'''

The DB is located in /data/mysql/data and the software apps are in /home/software/mysql/bin.  Access to both locations for user applications is defined in the module file.

See [[ MySQL v5 install ]] for installation details.