Alerting

Modern software applications are often complex, with many systems and services each of which has the potential to impact your users’ experience. This can make it difficult to tell when your applications are unhealthy and predict when problems are about to occur. With AppDynamics, you can set up detailed alerts that let you know when and where problems exist and also help you to anticipate problems that might be developing.  You can use these alerts to notify your operations teams and automate preventative actions to address the problems before they cause a severe slowdown or outage.

Overview

Out of the box, AppDynamics automatically discovers all the various components that make up your applications and how they interact.  Automatic baselining monitors performance metrics over time to determine what constitutes abnormal application performance for any given time and application load. You can then define health rules that will be triggered whenever slow or abnormal response times are detected or when thresholds for key metrics are exceeded.

AppDynamics has a few default Health Rules. For example, “CPU utilization is too high” triggers an event if the metric for CPU utilization exceeds a certain, defined threshold. You can also define your own Health Rules based on performance characteristics of your application. With a combination of these health rules and AppDynamics’ advanced statistical modeling that learns your application’s behavior, you can configure alerts that will proactively notify you when your application health is deteriorating or a rule has been violated.

AppDynamics’ Alert and Respond system is fully configurable to help you create customized alerts based on performance and availability metrics for your applications and the infrastructure on which they run.  Alerts are typically defined based on the following concepts, which each can be separately configured, and then associated with other elements to provide highly customized notifications:

  • Policies: Link health rule violations and other performance-based events with appropriate actions such as email notifications or diagnostic actions.

  • Health rules: Define key performance metrics for your applications across the entire platform and allow you to set thresholds for warnings and alerts.

  • Actions: Automate what should be done in response to health rule warnings or alerts, such as sending email/SMS notifications, running advanced diagnostics or recovery tasks.

  • Email digests: Send automatic email reports with details of specific events of particular interest whenever and wherever they occur.

Policies associate health rule violations and events to actions, so when an event corresponding to a policy is triggered or a health rule is violated, the policy causes a set of actions to be triggered based on the rules you set.  AppDynamics allows you to create highly granular health rules.  For example, you can create a health rule configured for just tier or perhaps for only certain nodes.

You can configure alerts on metrics gathered across the AppDynamics platform, so that you receive notifications whenever an error condition occurs, or slow business business transactions impact end user response times.  Alerts can be based on single, critical events, or performance trends over time that indicate underlying problems.  Here are some examples of things for which you might want to configure alerts

  • Business Transactions: business transaction response time is abnormally slow, either for a single transaction or average response over time

  • Application Runtime: application runtime metrics such as garbage collection or heap utilization show spikes or trends indicating application problems

  • Database performance: slow database queries or connection pool issues

  • Infrastructure performance: disk, CPU, memory or network performance metrics exceed set thresholds

  • Application errors and exceptions: configurable error conditions, such as runtime exceptions, HTTP status codes or method execution return values

  • Web and Mobile page load: slow loading pages, iFrames or Ajax requests

  • Custom metrics based on application-specific data

Based on the nature of the alert, you may want to configure remediation policies to ensure that trouble ticketing systems are updated, troubleshooting information is routed quickly to the right teams and, where appropriate, automated scripts are run to resolve the problem or provision additional resources.  Actions you can configure include:

  • running custom scripts or batch files to enable additional resources

  • making HTTP and REST service calls to trouble-ticketing systems

  • triggering advanced diagnostic actions to collect information about the problem

  • sending an email alert with a full digest of all the relevant metrics to your operations team

Setting up an alert is as simple as defining a policy, threshold, or health rule with the system or application metrics that you want to monitor.  After that you just need to define what action(s) you want to take in response.  These could range from an email or SMS notification, HTTP requests to provision additional resources, or integrations to applications like PagerDuty or ServiceNow.

The Policy Setup Wizard easily guides you through the alert setup process. 

Try AppDynamics

Introduction

Out of the box, AppDynamics automatically discovers all the various components that make up your applications and how they interact.  Automatic baselining monitors performance metrics over time to determine what constitutes abnormal application performance for any given time and application load. You can then define health rules that will be triggered whenever slow or abnormal response times are detected or when thresholds for key metrics are exceeded.

AppDynamics has a few default Health Rules. For example, “CPU utilization is too high” triggers an event if the metric for CPU utilization exceeds a certain, defined threshold. You can also define your own Health Rules based on performance characteristics of your application. With a combination of these health rules and AppDynamics’ advanced statistical modeling that learns your application’s behavior, you can configure alerts that will proactively notify you when your application health is deteriorating or a rule has been violated.

Understanding alerts

AppDynamics’ Alert and Respond system is fully configurable to help you create customized alerts based on performance and availability metrics for your applications and the infrastructure on which they run.  Alerts are typically defined based on the following concepts, which each can be separately configured, and then associated with other elements to provide highly customized notifications:

  • Policies: Link health rule violations and other performance-based events with appropriate actions such as email notifications or diagnostic actions.

  • Health rules: Define key performance metrics for your applications across the entire platform and allow you to set thresholds for warnings and alerts.

  • Actions: Automate what should be done in response to health rule warnings or alerts, such as sending email/SMS notifications, running advanced diagnostics or recovery tasks.

  • Email digests: Send automatic email reports with details of specific events of particular interest whenever and wherever they occur.

Policies associate health rule violations and events to actions, so when an event corresponding to a policy is triggered or a health rule is violated, the policy causes a set of actions to be triggered based on the rules you set.  AppDynamics allows you to create highly granular health rules.  For example, you can create a health rule configured for just tier or perhaps for only certain nodes.

What can AppDynamics alert to

You can configure alerts on metrics gathered across the AppDynamics platform, so that you receive notifications whenever an error condition occurs, or slow business business transactions impact end user response times.  Alerts can be based on single, critical events, or performance trends over time that indicate underlying problems.  Here are some examples of things for which you might want to configure alerts

  • Business Transactions: business transaction response time is abnormally slow, either for a single transaction or average response over time

  • Application Runtime: application runtime metrics such as garbage collection or heap utilization show spikes or trends indicating application problems

  • Database performance: slow database queries or connection pool issues

  • Infrastructure performance: disk, CPU, memory or network performance metrics exceed set thresholds

  • Application errors and exceptions: configurable error conditions, such as runtime exceptions, HTTP status codes or method execution return values

  • Web and Mobile page load: slow loading pages, iFrames or Ajax requests

  • Custom metrics based on application-specific data

What can I do with alerts

Based on the nature of the alert, you may want to configure remediation policies to ensure that trouble ticketing systems are updated, troubleshooting information is routed quickly to the right teams and, where appropriate, automated scripts are run to resolve the problem or provision additional resources.  Actions you can configure include:

  • running custom scripts or batch files to enable additional resources

  • making HTTP and REST service calls to trouble-ticketing systems

  • triggering advanced diagnostic actions to collect information about the problem

  • sending an email alert with a full digest of all the relevant metrics to your operations team

How do I set up an alert

Setting up an alert is as simple as defining a policy, threshold, or health rule with the system or application metrics that you want to monitor.  After that you just need to define what action(s) you want to take in response.  These could range from an email or SMS notification, HTTP requests to provision additional resources, or integrations to applications like PagerDuty or ServiceNow.

The Policy Setup Wizard easily guides you through the alert setup process. 

What's next

Try AppDynamics
Getting Started
  1. Install AppDynamics app agents on your system.

  2. Examine one of the default health rules. Click Alert & Respond > Health Rules.

  3. Create an alert. Click Alert & Respond > Actions > Create Action > Notifications > Send an Email.

  4. Create a policy that uses the examined health rule to trigger the alert. Click Alert & Respond > Policies > Create Policy. First define the TRIGGER, then the ACTIONS.

  5. If necessary, fine-tune the health rule to fulfill your needs. See Configure Health Rules.

Video Thumbnail
Advanced Training
Video Thumbnail
Virtual War Rooms