Steam Engine Cluster Administrator's Guide

Eric Nodwell

Scott Webster

This document can be freely redistributed according to the terms of the GNU General Public License.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.1 or any later version published by the Free Software Foundation; with no Invariant Sections, with no Front-Cover texts, and with no Back-Cover Texts. A copy of the license is included in the section entitled "GNU Free Documentation License".


Table of Contents

1. Administration
1.1. Adding a user
1.2. Monitoring the cluster
1.3. Changing wall times or other scheduler parameters
1.3.1. Changing the default wall time
1.3.2. Changing one user's wall time
1.4. Reinstalling a node
1.5. Swapping Nodes
1.6. Changing the Number of Nodes
1.7. Backing up the master node
2. Master Node Installation
2.1. Basic Debian installation
2.2. Some networking stuff
2.3. inetd
2.4. ssh
2.5. rsh
2.6. postfix
2.7. ntp
2.8. Printing
2.9. Firewall
2.10. Create a local Debian mirror
2.11. Back-port certain packages from Testing
2.12. Install 2.4 series kernel
2.13. lm-sensors (CPU temperature monitor)
2.14. Channel-bonding
2.15. Configure update-cluster
2.16. C3 Cluster Command and Control Tools
2.17. Matlab
2.18. Lahey Fortran
2.19. MPI-CH
2.19.1. BLACS
2.19.2. ScaLAPACK
2.20. Java
2.21. mysql
2.22. Maui Scheduler
3. Node installation with FAI (Fully Automatic Installation)
3.1. Download and install FAI
3.2. Basic configuration
3.3. BIOS configuration on the nodes
3.4. bootp
3.5. Target configuration
3.5.1. /var/fai/config/class
3.5.2. /var/fai/config/disk-config
3.5.3. /var/fai/config/package_config
3.5.4. /var/fai/config/scripts
3.5.5. /var/fai/config/files
3.5.6. /var/fai/config/hooks
3.6. Installing one or more nodes