DESCRIPTION

Sun Grid Engine is a facility for executing UNIX batch jobs (shell scripts) on a pool of cooperating workstations. Jobs are queued and executed remotely on workstations at times when those workstations would otherwise be idle or only lightly loaded. The work load is distributed among the workstations in the cluster corresponding to the load situation of each machine and the resource requirements of the jobs.

User level checkpointing programs are supported and a transparent checkpointing mechanism is provided (see Checkpointing jobs migrate from workstation to workstation without user intervention on load demand. In addition to batch jobs, interactive jobs and parallel jobs can also be submitted to Sun Grid Engine.

USER INTERFACE

The Sun Grid Engine user interface consists of several programs which are described separately.

qacct extracts arbitrary accounting information from the cluster logfile.

qalter changes the characteristics of already submitted jobs.

qconf provides the user interface for configuring, modifying, deleting and querying queues and the cluster configuration.

qdel provides the means for a user/operator/manager to cancel jobs.

qhold holds back submitted jobs from execution.

qhost displays status information about Sun Grid Engine execution hosts.

qlogin initiates a telnet or similar login session with automatic selection of a low loaded and suitable host.

qmake is a replacement for the standard Unix make facility. It extends make by its ability to distribute independent make steps across a cluster of suitable machines.

qmod allows the owner(s) of a queue to suspend and enable all queues associated with his machine (all currently active processes in this queue are also signaled) or to suspend and enable jobs executing in the owned queues.

qmon provides a Motif command interface to all Sun Grid Engine functions. The status of all or a private selection of the configured queues is displayed on-line by changing colors at corresponding queue icons.

qquota provides a status listing of all currently used resource quotas (see

qresub creates new jobs by copying currently running or pending jobs.

qrls releases holds from jobs previously assigned to them e.g. via (see above).

qrdel provides the means to cancel advance reservations.

qrsh can be used for various purposes such as providing remote execution of interactive applications via Sun Grid Engine comparable to the standard Unix facility rsh, to allow for the submission of batch jobs which, upon execution, support terminal I/O (standard/error output and standard input) and terminal control, to provide a batch job submission client which remains active until the job has finished or to allow for the Sun Grid Engine-controlled remote execution of the tasks of parallel jobs.

qrstat provides a status listing of all advance reservations in the cluster.

qrsub is the user interface for submitting a advance reservation to Sun Grid Engine.

qselect prints a list of queue names corresponding to specified selection criteria. The output of qselect is usually fed into other Sun Grid Engine commands to apply actions on a selected set of queues.

qsh opens an interactive shell (in an on a low loaded host. Any kind of interactive jobs can be run in this shell.

qstat provides a status listing of all jobs and queues associated with the cluster.

qtcsh is a fully compatible replacement for the widely known and used Unix C-Shell (csh) derivative tcsh. It provides a command-shell with the extension of transparently distributing execution of designated applications to suitable and lightly loaded hosts via Sun Grid Engine.

qsub is the user interface for submitting a job to Sun Grid Engine.

COPYRIGHT

Copyright: 2008 by Sun Microsystems, Inc.