ACT Utils is a part of Advanced Clustering's ACT dir software installation. Read more about everything included with ACT dir.
Advanced Clustering's ACT Utils package provides an easy-to-use command-line-based set of management utilities for your Apex Compute Cluster. These powerful parallel shell tools can be included as part of scripts or used interactively as part of daily administrative tasks.
- Consistent command-line options across all utilities
- Easy INI file syntax that allows you to specify all nodes in a cluster in just a few lines
- Specify individual nodes, multiple nodes, or pre-defined groups
- Easy custom-defined grouping of nodes so you can easily act on subsets of nodes (common group examples: cabinet1, cabinet2, ibnodes, etc.)
- Commands run in parallel so they work quickly and efficiently
- Uses existing remote execution software already in place — no new server or daemons to run (configurable between SSH and RSH).
Below you find some of the commands that are included in the ACT Utils package and some examples for how to use the commands. If you have an Apex Cluster and need assistance with ACT Utils, contact our support department for assistance.
act_exec — Run any command in parallel on the nodes you specify.
Example: To get the load average and time the system has been up on a group of nodes called cabinet1.
[root@testcluster ~]# act_exec -g cabinet1 uptime
10:02:53 up 4 days, 3:56, 0 users, load average: 0.07, 0.02, 0.00
10:02:53 up 4 days, 3:55, 0 users, load average: 0.10, 0.04, 0.00
10:02:53 up 4 days, 3:55, 0 users, load average: 0.02, 0.06, 0.00
10:02:53 up 4 days, 3:56, 0 users, load average: 0.08, 0.03, 0.00
act_cp — Copy a file from the machine you are logged into to the nodes you specify.
Example: To copy the file "bigfile" from the act user's home directory to /tmp on nodes01 and node04.
[root@testcluster ~]# act_cp -n node01,node04 ~/act/bigfile /tmp/bigfile
act_ipmi — Interface to IPMI (Intelligent Platform Management Interface) cards on nodes. Supports power control (power on, off, reboot and status), fan RPMs, system temperatures, voltages, and remote console.
Example: To get the system temperatures, voltages, and power status of node01 and node04.
[root@testcluster ~]# act_ipmi -n node01,node04 temps
[root@testcluster ~]# act_ipmi -n node01,node04 voltages
[root@testcluster ~]# act_ipmi -n node01,node04 power-status
act_killall — Kill off a process by name on a group of nodes you specify.
Example: To kill the process "myjob" on all machines in the cluster.
[root@testcluster ~]# act_killall -a myjob
act_shutdown — Shut down, power off, or reboot nodes or subsets of nodes.
act_rm — Remove files from nodes or subsets of nodes.
act_authsync — Synchronize all authentication database files between nodes in your cluster.
act_power — Interact with network controlled power distribution units. Hardware power on, off, or reboot individual nodes in your cluster.