Managing Application Templates

Prerequisite: You must have “VSE All Tools” authorization to modify template attributes. (See “User Authorizations” in VSE Management Software Help for a discussion of VSE tool box permissions.)

Find an Application Template

You can access individual application templates by clicking the AD Templates tab, then the Admin/Config tab.

In the Templates table, click the name of the template that you want to modify.

The data for that template opens in a new screen.

TIP:

Another way to get to a template for editing is to click on the associated application name in an application table listing. When the information for the application is displayed, you can see a link to the template for that application in the Application Summary box.

Example template
link on application detail screen

Change AD Application Template Values

To edit the selected application template, do the following:

Click the Edit button to reveal the fields that can be edited.
Change the field values as desired to describe the processes that constitute this application.
Click Apply to overwrite the template with the new values. (You might need to scroll to the right to see the Apply and Cancel buttons.)

Once you apply the application template, a short time passes while Application Discovery attempts to match the newly edited application to discovered processes currently running in the network. As matching completes, applications matching this template description appear in the table called 'Applications effected by this template' (below the template summary boxes).




	NOTE: My application does not appear in the "Applications effected..." table. Why? There are several possible explanations for this to occur: It takes some time for template matching to complete. Action: Wait a couple of minutes to see whether or not matching occurs. For a more precise indicator of the matching interval (and therefore, the likely wait-time), click the Admin/Config tab, then the AD Agents tab, and find the column labeled 'Proc intrvl' (Process interval) for the server where you expect the application to be running. The process interval roughly represents the time needed to refresh the matching data. There is an error in the template values such that the application described in the template does not match any currently running processes. Action: Double-check your template values and edit if necessary. The template is correct, but the application described is not currently running. Action: Ensure that the application designated in the template is running on the host where you expect it to be. For more information, see “Data Missing in Application Discovery Screens” in the troubleshooting section of Application Discovery Help.

Create a New Template

The process for creating a new application template is similar to editing an existing template.

To start, click the Admin/Config tab, then the AD Templates tab.
Click the Create New AD Template button.
The application template editor opens.
Fill in the application template values as desired. See “Key to Application Template Values”.
Apply the changes. This action saves the new template under the unique name that you have supplied. (You might need to scroll to the right to locate the Apply and Cancel buttons.)

Your new application template immediately appears in the list of all templates that is displayed when you click the Admin/Config tab, then the AD Templates tab. The new template definition is applied the next time that matching is compared to the collected data.

Key to Application Template Values

The tables on this page provide the definitions for the editable fields in HP Application Discovery (AD) templates.

Application Identity and System Scope Fields These fields let you provide unique identifiers for the template itself. These fields also include options for restricting the application definition to a particular operating environment.

Separation Rule Fields These fields let you set limits on how many users and how many processes can be associated with the template. You can also set visibility of the resulting application in HP Application Discovery screens.

Aggregation Rule Fields These fields let you build a rule that describes the process constituents for the resulting application. Regular expressions are allowed in these fields when designating user, executable path, and command line arguments.




	Constructing Regular Expressions: HP Application Discovery recognizes regular expressions constructed using Perl 5 or POSIX syntax and semantics. To learn more, consult PCRE - Perl Compatible Regular Expressions at http://www.pcre.org/.

NOTE:

HP Application Discovery (AD) interprets executable path strings differently than HP Global Workload Manager (gWLM).

HP Application Discovery looks at a process from the perspective of how it is named when it is running. Therefore, to successfully construct an application that HP Application Discovery recognizes, you must provide attributes (data strings) that describe the running process.

Also, when defining an application, keep in mind that HP Application Discovery looks across your entire network for likely matches, and not just on one particular managed node.

It might help to think of the difference between AD and gWLM this way:

In AD, an executable and arg[0] are compared directly to one another to make a match.
In gWLM, the inode of an executable and the inode of arg[0] are compared to one another to make a match.

The following examples demonstrate the different results that you can expect when using the same executable and arguments in HP Application Discovery as compared to HP Global Workload Manager (gWLM).

Table 3-3 An Example Comparison of Match Behavior in AD and gWLM for Same Executable Path and Arguments

Example Inputs		AD	gWLM
Executable	Process arg[0]	AD	gWLM
`/usr/sbin/biod`	`/usr/sbin/biod`	match	match
	`biod`	no match	match
	`./biod`	no match	match
`biod`	`/usr/sbin/biod`	match	not allowed
	`biod`	match	not allowed
	`./biod`	match	not allowed

Application Identity and System Scope Fields and Their Descriptions

Field	Description
Application Identity
`Name`	Required. Unique name given to the template to represent the application.
`Version`	Required. Version number of the application this template matches.
`Title`	User-defined name for the template.
`Description`	User-defined summary describing the template.
System Scope
`OS`	Operating system to which the template applies. If this field is left blank, HP Application Discovery tries to match the template within all operating systems for which it has information.
`OSver`	Operating system version to which the template applies. If this field is left blank, HP Application Discovery tries to match the template against all operating system versions for which it has information.
`Architecture`	Architecture to which the template applies. If this field is left blank, HP Application Discovery tries to match the template against all architectures for which it has information.




	NOTE: If you want to define a template based on where the processes that you want to collect into an application are running, you can check the 'Unmatched processes' table on a particular host screen or the 'Processes not matched by this template' table on a template editing screen to get this information. See Filling in the System Scope Fields Using Table Data for tips on finding specific process attributes for this purpose.

Separation Rule Fields and Their Descriptions

Field Description

Is visible?

By default during initial discovery, HP Application Discovery sets objects recognized as packages to "no" and objects recognized as templates to "yes".

This field designates whether or not the application shall be visible to users of the HP Application Discovery user interface:

yes – show the application data associated with this object.
no – hide the application data associated with this object.




	NOTE: Visibility actually has no affect on template matching, but is included here as a convenience. It is a global setting that allows you to control data presentation in all HP Application Discovery screens. It is not stored as a user preference.

Pid scope

How many process identifiers (pids) are allowed in an application:

one — one pid per running application.
Effect: Each new concurrent run of this application is shown as a separate instance from the others in HP Application Discovery.
session – any number of pids per application for the duration of one terminal group session.
Effect: All associated pids are treated as one running application instance for that session. However, because sessions on HP-UX and Linux are maintained per user, this has the effect of delineating running applications per user. Thus, if you log in twice (two concurrent terminal windows), you can have two separate instances of this running application. This effect is similar to that obtained by setting user scope=one.
any – any number of pids per application.
Effect: Pids have no specific effect on defining the application described by this template.

User scope

How many users' processes can be associated with one instance of this application:

one – one user's processes per application.
Effect: A separate instance of this application will be shown for each different user who runs this application. This effect is similar to that obtained by setting pid scope=session.
any – any number of different users per application.
Effect: Number of users has no specific limiting effect on defining the application described by this template. That is, only one instance is recognized, regardless of how many users are running the application concurrently.

is pidtree

Shall child processes spawned by the parent be considered part of the parent application instance or as a separate instance:

no – child processes are not considered part of the parent application instance. This is the default value.
Effect: Child processes will be shown as separate application instances in HP Application Discovery.
yes – child processes are considered part of the parent application instance.
Effect: Child processes will not be shown as separate application instances in HP Application Discovery.




	NOTE: If you want to define a template based on ownership or how many instances are currently running, you can check the 'Unmatched processes' table on a particular host screen or the 'Processes not matched by this template' table on a template editing screen to get this information. See Filling in the Template Separation Rule Fields for tips on finding specific process attributes.

Aggregation Rule Fields and Their Descriptions

Field Description

Rule

Rules can be simple or complex, and are based on types that can be added to or deleted from the rule. Possible type values used to construct a rule include:

Or
Can contain a list of one or more And, Or, User, Group, Path or Cmdline values.
Effect: Any one of the rules within this statement can be true to match a process to the application.
And
Can contain a list of one or more And, Or, User, Group, Path or Cmdline values.
Effect: All of the rules within this statement must be true to match a process to the application.
User
Supply the user name associated with the process. In addition to literal names, regular expressions are allowed.
Effect: Make matching to a process dependent on a particular user.
Group
Supply the group name of the process. Regular expressions are allowed.
Effect: Make matching to a process dependent on a particular group.
Path
Supply the executable path of the process. Regular expressions are allowed.
Effect: Make matching to a process dependent on a particular executable path.
Cmdline
Supply the command line arguments used to start the process. Regular expressions are allowed.
Effect: Make matching to a process dependent on a particular command line argument.




	NOTE:

For an example of how to build an application template rule for aggregating processes into an application, see Create a Process Aggregation Rule.

Create an Application Template Rule

Application Discovery provides a set of options and fields that help you to define a valid application template rule. The rule you create governs discovery of the application that you have defined. You can create rules that apply generally or specifically to a process, user, or group.

Create an Aggregation Rule

An aggregation rule is one in which you designate one or more characteristics in order to collect the associated process or processes into an application that can be matched by Application Discovery.

The following procedure assumes that you have opened an application template and are ready to edit the Rules form.

To start, select a radio button for the first type field that you want to add and click the Add Rule button. In the example below, the first type added is an And operator.
To add a type within the And statement braces, select a type, and click the Add Rule button aligned with the dot (.) within the And statement. In the example below, a User type is added when the button is clicked.
The result when the Add Rule button is clicked is as follows:
You can add a new type after "User" within the original And statement by clicking the Add Rule button at (below), or you can add additional statements or types after the original And statement by clicking the Add Rule button at .
Continue in this way until you have added all the types that you want in the rule.
Fill in the empty values to complete the rule description for this template, and click Apply. (You might need to scroll to the right to locate the Apply and Cancel buttons.) For guidance on the values that work in each field, see “Key to Application Template Values” and “Check Unmatched Processes to Find Application Candidates”.

Aggregation Rule Examples

The following images demonstrate several ways to construct an aggregation rule in Application Discovery that will collect exactly the processes that you want associated to this template to be defined as an application.

The following figure shows a simple aggregation rule that collects processes by specified executable path and command line argument.

Figure 3-6 Example of a Simple Template Aggregation Rule Showing Executable Path and Argument

The following figure shows an aggregation of multiple processes by specified executable path into one application.

Figure 3-7 Example of a Template Aggregation Rule that Collects Multiple Processes

The following figure shows an aggregation of multiple processes specified using regular expressions.

Figure 3-8 Example of a Template Aggregation Rule that Uses Regular Expressions

How to Remove Types from the Aggregation Rule

You might want to remove types from a rule to correct it or refine it. To remove a type, click the Del Rule button to the right of the type field that you want to remove. To delete And or Or statements, click the Del AND or the Del OR button to the right of the closing brace for that statement. When an And or an Or rule is deleted, all the rules that it encloses are deleted also.

This example shows which Del Rule button to click to remove the Group type within an And statement and to remove the entire And statement.

Figure 3-9 Example Demonstrating the Location of the Del Rule Button in Relation to its Rule

Check Unmatched Processes to Find Application Candidates

In Application Discovery, processes running on a host system that do not match existing template or package definitions are called unmatched processes. Characteristics of unmatched processes can be entered into a template to define an application.

Understanding the Tables that Show Unmatched Processes

The create and edit application template screens show unmatched processes found across the computing network in a table called "Processes not matched by this template". The same data types are also displayed for a particular host on each individual host page in a table called “Unmatched Processes”.

Figure 3-10 Example “Unmatched Processes” Table

You can sort or filter this table to find processes with which you want to define an application.




	TIP: One way to reduce the amount of 'noise' cluttering your tables is to create a template that collects 'uninteresting' processes into an application that can be hidden.

Filling in the System Scope Fields Using Table Data

In the following image, you can see the columns directly relevant to the System Scope fields in an application template (dark gray column headings). By entering a specific operating system (OS), operating system version (OSver), and/or architecture (Arch), you can limit the application of the template to particular systems in your network.

Figure 3-11 Finding Data to Enter in System Scope Area of Application Tempate

For more details about this part of an application template, see “Application Identity and System Scope Fields and Their Descriptions”.

Filling in the Template Separation Rule Fields

In the following image, you can see the columns indirectly relevant to the Separation Rules fields in an application template – pid, ppid, user, sid, and ct. These columns are indirectly relevant because you do not use the actual values in the columns to define the separation rule. Instead, this information is provided to help you decide which attributes are appropriate to select.

For example, when viewing the data in the table, look at the relationships among process IDs, users, and sessions to determine what relationships you want to define in the Separation Rules.

Figure 3-12 Finding Data to Enter into Separation Rule Fields

For details about setting attributes in an application template, see “Separation Rule Fields and Their Descriptions”.

Filling in the Aggregation Rule Fields Using Table Data

In the following image, you can see the columns directly relevant to the Aggregation Rule fields in an application template. By entering one or more users (User), groups (Group), paths (Path), and/or arguments (Cmdline), you can limit the application of the template to particular processes running in your network.

Figure 3-13 Finding Data to Enter into Aggregation Rule Fields

For more details about this part of an application template, see “Aggregation Rule Fields and Their Descriptions” and “Create an Aggregation Rule”.

Create an Application Workload in Virtualization Manager

The following procedure illustrates the case where an application manager wants to create an application-based workload using processes discovered by Application Discovery, but that are not currently recognized as an application by either Application Discovery or Visualization Manager.

Prerequisites:

You must have “VSE All Tools” authorization to modify template attributes. (See “User Authorizations” in VSE Management Software Help for a discussion of VSE tool box permissions for Application Discovery and Virtualization Manager.)
The processes that you want to include in the workload must be running on the system where you are creating the workload.

Start by creating a workload in Visualization Manager.
To get to Virtualization Manager, select ToolsVSE Management... on the HP SIM menu bar.
The Virtualization Manager Visualization
tab opens.
Click CreateWorkload Definition... on the Visualization tab menu bar.
The Create Monitored Workload Definition screen opens.
Choose a system, and type a workload name and description.
Specify the Workload Definition Criteria to be “by application” in the drop down.
Optionally, click “Show All Running Applications” to see all currently discovered applications on the selected system.
Click the Create/Edit Application Template... link.
The AD Templates tab opens in the Application Discovery user interface.
Click Create new AD template.
A new application template opens for editing.
Fill in the Application Identity, the System Scope, the Separation Rules, and the Aggregation Rules.
Click Submit.
Click OK in the dialog box to confirm the template submission.
The new template displays.
Optionally, click the Admin/ConfigAD Agents tab to see what the process interval is for the target system of this application template.
Assuming that the process interval is 5 minutes, that is the length of time to wait until Application Discovery finds the new template to match to running processes on the target system.
Click the link “Go back to Create Monitored Workload Definition” that appears under Application Discovery at the top of whatever screen you are on.
The Create Monitored Workload Definition screen reopens.
Check the table labeled Application Name and Version for your new application.
Click the radio button to the left of your new application's name.
Click Add to Definition.
Optionally, click the check box to include all versions of the selected applications in this workload.
Click OK.
The Virtualization Manager Workload tab screen opens with your new application-based workload included in the list.