Introduction to Application Monitoring & Control

From OpsWise Documentation Wiki

Contents 1 Introduction 2 Processing Flow

Introduction

The OpsWise Application Monitoring and Control features allow you to use OpsWise Automation Center as a network control and monitoring tool. You can use these resources to start, stop, and query any application running on any machine where you have an OpsWise agent installed and running.

You will use the following three components to monitor your applications:

• The Application Resource record defines the name and location of the application, along with the necessary control commands. The list of application records displays a status for each application.
• Three Application Control tasks are automatically generated when you save an Application Resource record to the OpsWise database: one each for executing a start, stop, and query command against the application. You can also create customized Application Control tasks where necessary.
• Optional Application Monitor Triggers allow you launch other tasks based on the status of the Application Resource record being monitored.

Processing Flow

Described below is the processing flow that occurs when you set up an application to be monitored by OpsWise Automation Center.

1. For each application you want to monitor, you use the Application Resource screen to create an Application resource record. This record provides the information necessary for OpsWise to start, stop, and query the application.
   The Application resource is used in conjunction with Application Control tasks. Each time you create and submit an Application resource record to the database, OpsWise automatically creates three related tasks, one each for starting, stopping, and querying the application. Although you can also execute these tasks as you would any other OpsWise task, in this context the tasks work differently, as follows:
   • You will execute them from the Application list using the drop-down or right-click menu, as described in the next step.
   • When you execute them from the Application list, OpsWise does not create a task instance record. It simply starts, stops, or queries the application. In this context, no output is gathered and saved to the database. However, if you execute the tasks using triggers, workflows, or by manually launching them from the task list, they execute like any other task; that is, the system creates a task instance, the task appears in the Activity screen, and any output collected is attached to the record.
2. Once you have added the Application resource record(s), you start the application using one of the following methods:
   1. Click the box to the left of the Application name to select it. You can select as many Applications as you want.
   2. From the Actions on selected rows... menu, select Start.
   3. 

   Or

   1. To start a single application, right-click the Application Name and select Start from the pop-up menu.

   

3. The system executes the Start command, as provided by the user in the Application Resource definition. It puts the Application into Starting status, and saves the Start Time. The Start command has two functions: 1) it starts the application and 2) it starts the query process that monitors the application.
4. After 30 seconds, the system automatically executes a Query command, as provided by the user in the Application resource definition, to determine the status of the application. The system continues executing the Query command every 120 seconds thereafter until the user stops the monitoring by issuing a Stop command from OpsWise.



5. The purpose of the Query is to determine whether or not the Application is Active. The system uses the specifications provided by the user in the Exit Code Processing fields (example shown below) to make this determination.



• If the response from the application indicates a successful startup, the system puts the Application in Active status.
• If the response indicates the Application hasn't started, the system continues executing the Query (keeping track in the Startup Query Attempts field) until it reaches the maximum attempts specified by the user in the Startup Query Maximum field. If the maximum number is reached before achieving an Active status, the system puts the Application into Impaired status. However, the system continues monitoring. If the appropriate exit code parameters are eventually returned, the application will go to Active status. The purpose of the Startup Query Attempts field is to avoid having the application go right into Impaired status if it takes awhile to start.

The system writes any Exit Code captured by the Query in the Query Exit Code field.

6. After starting the application, OpsWise continues monitoring by sending out the Query commands every 120 seconds.
   • If OpsWise detects a problem based on the Exit Code parameters, it puts the Application into Impaired status. If this occurs, you have several options for handling the problem, with increasing levels of automation:
     • The Application list displays the status of all Applications being monitored. You could create a filter for the Application list that displays only those with status of Impaired (or other), as shown in the example below. When you see a problem, troubleshoot the issue and restart the Application from outside OpsWise.
     
     • Set up an Application Monitor trigger that monitors this application for Impaired and other problem statuses. When the trigger is satisfied, it launches an email task that sends emails to support personnel, notifying them of the problem. Several built-in variables are supported that allow you to pass required data into the email message: the application name, type, and status.
     • You could also create a workflow launched by an Application Monitor trigger looking for Impaired or other problem statuses. The workflow can include Application Control tasks that attempt to resolve the problem by stopping and then re-starting the application. You could also include any other tasks that are specific to troubleshooting the application.
   • If OpsWise fails to get a response to a Query for three minutes, it puts the Application into <status>/Query Overdue status, where <status> is the last known query status. This can be any of the following: Starting/Query Overdue, Active/Query Overdue, and Impaired/Query Overdue. For example, you may see this status if the agent went down or there was some other problem on the machine unrelated to the application itself. If this occurs, you should troubleshoot the issue. When you have fixed the problem, the continued queries from OpsWise will then return an Active status for the application.
7. To stop monitoring an Application, issue the Stop command against it. OpsWise stops the Application and puts it into Inactive status, which means it is no longer monitoring.