|Caldera Volution Manager Administration Guide|
Using Caldera Volution Manager you can combine and customize components to automate complex tasks. This topic provides several examples, including:
Section 6.1 - working with computer groups and profiles to automate software updates to specific types of systems
Section 6.2 - constraining actions to affect only targeted systems
Section 6.3 - creating custom actions to automatically respond to system problems
Keeping software up to date is a time-consuming task. Updates need to occur regularly as:
popular industry-standard applications, such as web browsers, are updated to make use of the latest technology
security alerts are issued that require timely updates of systems containing vulnerable packages
your company decides to deploy a new tool or application
You might spend many hours creating reference platforms and distributing media to your systems for installation. Or, perhaps, you update your systems over the network using scripts and manual processes.
You can use VM to automate these tasks:
Populate the Software Repository package distribution directories with the packages you want to distribute. (See Section 3.1.1).
Create computer groups containing systems with a similar purpose. For example, you might create a group for departmental servers, one for web servers, one for database servers, and one for user workstations.
You can create these groups manually or automatically (based on a hardware or software inventory). For example, if you know that the database servers on your network are the only ones currently installed with a particular database software package, you can use the Software Inventory action to scan your network, then create a group based on those systems. For information about creating computer groups, see Section 4.4.
Create profiles matching the groups you created. For example, the workstations profile might contain personal productivity applications, web and multi-media tools, e-mail, and other packages that the web server profile does not include. See Section 3.2.
Once your groups and profiles are defined, software updates on your network are greatly simplified. All you need to do is update the repository and the desired profiles.
Many corporations have a mixed computing environment. Due to hardware from different vendors, corporate mergers, or even organizational preferences, it is possible that your network might contain computers running various brands and releases of UNIX systems and Linux systems.
VM helps you to target specific computers to update through the use of constraints. We previously described how to update several groups of computers using profiles tailored to their purpose - if all systems in the web servers group run a single operating system platform (for example, Caldera OpenLinux Server 3.1.1), a single profile can be implemented that installs the right software for that operating system. But what if the web servers group contained systems running SCO OpenServer 5.0.6 and Open UNIX 8 as well? You have two choices:
You can further break down your computer groups into OS-specific groups (for example, SCO OpenServer web servers, Open UNIX 8 web servers, and OpenLinux web servers). You can then create and link a specific profile to each group. This works, but it is time-consuming and leads to a cluttered directory structure, with possibly more groups than desired.
The alternative is to leave the original computer groups intact, but create three web server profiles, each constrained to run only on the systems matching the specified operating system. You can then link all three of these profiles to the web servers group, and VM automatically determines which profile is run on each client. See Section 18.104.22.168 for more information.
The ability to constrain actions to particular types of systems lets you to simplify your group structure and target only the computers you want to affect. It is useful not only for software distribution, but when performing any type of action, including custom actions.
The Health Monitor policies let you automatically detect shortages of memory, disk space, processes, and other resources. This is a highly-configurable feature. You can:
define thresholds for notices, alerts, and warnings for each item
define how these notices, alerts, and warnings are communicated by defining the Gateway Policy
define custom actions to be automatically run each time a defined threshold is exceeded
Because each parameter, such as free disk space, can be individually configured, and custom actions you define are run at thresholds you indicate, there are a large number of possible implementations. Here are some examples of what you can do:
Create a custom action so that whenever the free disk space notice threshold is exceeded, a warning e-mail is sent to users notifying them to clean up their temp files. A separate action triggered by the warning condition could automatically remove all temp and core files from the system.
Create a custom action so that, if the load average warning threshold is exceeded, the process table is scanned and runaway processes are killed.
Create a custom action so that, when the number of unique users logged on warning threshold is exceeded, the number of users and time are logged. You might use this information to load-balance users across a number of systems.
Create a custom action so that, if the memory hog notice threshold is exceeded, information about the offending process and user is logged. You could use this information to identify software or users that need to move to a system with more memory, or to justify hardware upgrades.
These actions might be called from a single Health Monitor policy. Or, you might create multiple policies, each linked to different computer groups or constrained to run on only a certain operating system or vendor platform. For more information, see Section 4.5.