| Scyld ClusterWare HPC: Administrator's Guide | ||
|---|---|---|
| << Previous | Next >> | |
Included in the Scyld ClusterWare distribution are extra tools that may be of interest to users, including the TORQUE resource manager, IPMITool for monitoring and managing compute node hardware, and the Ganglia monitoring tool.
TORQUE is an open source tool based on standard OpenPBS, providing users with an intuitive interface for remotely initiating and managing batch jobs on distributed compute nodes.
The TORQUE resource manager is included in the default installation for Scyld ClusterWare HPC. TORQUE is not included in the default installation of Scyld Beowulf Series 30.
The TORQUE daemons on the master node are enabled and started by default when the master node boots up. However, following the first install of ClusterWare TORQUE, or when the Release Notes for a ClusterWare upgrade instructs you, TORQUE requires running these additional commands:
beochkconfig 90torque on to enable the TORQUE startup script that executes on every compute node at cluster startup time.
service torque reconfigure to direct TORQUE to self-configure. NOTE: this reconfigure resets to default values all existing TORQUE configuration settings and databases.
service torque start to restart TORQUE on the master node.
bpctl -S all -R to reboot the compute nodes. TORQUE is now available across the cluster.
You can refer to the following sources for additional information on TORQUE:
For basic information on using TORQUE, see the Scyld ClusterWare HPC User's Guide.
For more general product information, see the TORQUE information page sponsored by Cluster Resources, Inc. (CRI).
| << Previous | Home | Next >> |
| Implementing a Scheduling Policy | IPMITool |