This article link: http://blog.csdn.net/kongxx/article/details/8022496
First from http://www.adaptivecomputing.com/resources/downloads/torque/the latest installation package, which uses the version 4.1.2, the filename is torque-4.1.2.tar.gz.
Installation
$/configure
$ make
$ sudo made install
$ make packages (used to generate installation packages for installation on other machines, resulting in several torque-package-*.sh files after running )
After installation, view the environment variable "path" to ensure that/usr/local/bin and/usr/local/sbin are in "path". Switch to root user
# echo '/usr/local/lib ' >/etc/ld.so.conf.d/torque.conf
# ldconfig
Initialization and Configuration
#./torque.setup <user>
Start stop service
Here it is important to note that the torque is made up of three main components
Pbs_server PBS Service Daemon, responsible for receiving job submissions, located on service node
Pbs_sched PBS Dispatch Daemon, responsible for scheduling jobs, located on service nodes
Pbs_mom PBS Mom Daemon, responsible for monitoring native and performing jobs, on all compute nodes
You can use the following command to start the torque service
$ sudo pbs_server
$ sudo pbs_sched
$ sudo pbs_mom
$ sudo trqauthd
In addition, torque also provides scripts that allow you to use Linux service to start stop torque by copying the following files from the torque directory to the/ETC/INIT.D directory, as follows:
$ sudo cp contrib/init.d/pbs_server/etc/init.d/
$ sudo cp contrib/init.d/pbs_sched/etc/init.d/
$ sudo cp contri b/init.d/pbs_mom/etc/init.d/
$ sudo cp contrib/init.d/trqauthd/etc/init.d/
You can then use the following command to start the stop service
$ sudo/etc/init.d/pbs_server start|stop|status
$ sudo/etc/init.d/pbs_sched start|stop|status
$ sudo/etc/ Init.d/pbs_mom start|stop|status
$ sudo/etc/init.d/trqauthd start|stop|status
Test
1. Write a test script/opt/test/sleep.sh on the master machine, which reads as follows
#!/bin/sh Sleep
600
2. Submit Homework
$ qsub/opt/test/sleep.sh
3. Check Homework
$ qstat-a-N
You can see the running status of each job at this time.