How do I set up an alert for when an AppDynamics agent is down?

[#1]

What are the steps for setting up an alert to be notified when an AppDynamics agent is down?

There are three steps to creating an agent-down alert. You will need to set a health rule, create an action, and then create a policy.

Contents

1. Set a health rule
2. Create an action to send an email when the Agent is down
3. Create a policy and select the health rule on which the policy should be evaluated

---

 

^{STEP 1} Set a health rule

To receive an email alert when an app agent or machine agent is down, you will need to set up a health rule from the Controller UI. 

1. Open Create Health Rule
   
   
   
2. Set the health rule to affect application performance.
   
   
   
3. Input x in "Trigger only when violation occurs x times in the last 10 min(s)"
   

Note: For machine agent availability, select the availability metric under Agent > machine.

^{STEP 2} Create an action to send an email

Next, create the action to send an email when the agent is down. 

1. From the Controller UI, create a new action.
   
   
   
2. Enter the email address that should receive the alert.
   

^{STEP 3} Create a policy and select the health rule

Finally, create a new policy and select the health rule you created in STEP 1.

1. From the Controller IU, create a new policy.
   
   
2. Select the email address.
   

Additional resources:

For instructions on how to set up alerts to monitor the health status of your application, please visit these resources:

• Health Rules
• Configuring Health Rules

[Abhilash_C]

Hi ,

What if the Application Don't get Sufficent load for last 30 minutes ? Still App agent woud be Going down right ? How to overcome this scenario ?

[CommunityUser]

App agent is down we create the health rule from the above procedure, Can we use the same health rule for multiple Applications?

[Vijayakumar_Gov]

@Abhilash.C, @Anonymous ,

The above mentioned procedure will trigger the alert even if the agent goes down and not reporting (Evaluate to true on no data checkbox should be checked). All the HRs will be created under specific application only. 

Thanks,

Vijay

[Anand_Shibaraga]

HI Vijay,

Why there is a  difference when its app agent SUM < 5 and machine SUM>5. Also, why do we use a custom metric, cant we just use node health option?

Thanks

Anand

[Vijayakumar_Gov]

Hi @Anand.Shibaragatti ,

For both App agent and Machine agent, SUM should be less than the respective value. There is no difference in the availability metric for agents.

Custom Metric shown here is for representation purpose only. This method works fine, if you select availability metric in any other way.

Thanks,

Vijay

[Anand_Shibaraga]

Hi Vijay,

Thanks I used the same SUM  < specific value 5 for last 10 mins.  Hope this works. 

Any idea how this can be used in cloud where the servers are dynamic. 

Thanks

Anand

[Claudia_Landiva]

Hi, @Anand.Shibaragatti

Are you in a position to share how using the SUM <specific value 5 for 10 minutes worked out for you? Hope things are progressing.

Claudia Landivar

Communuty Manager & Editor

[Miguel_JimeneZ]

Hello

They will have an update on how to perform the procedure with the new App D updates .... I am looking for a way to know when the agent crashes.

[Claudia_Landiva]

Hello, there @Miguel.JimeneZ 

We're definitely assessing this article for revision: thank you for calling it out. I will keep everyone posted here.

In connecting with the Alerting PM, he confirmed that the Availability metric is remains usable in Health Rules. He might need to learn more about what you're trying to accomplish, and the issues you're facing. Could you elaborate here? Of course, if logs or ppi are involved, you would need to open a Support ticket.

One thing which has changed is the deprecation of the SUM feature. Instead, users should input x in "Trigger only when violation occurs x times in the last 10 min(s)". If you try that, will you let us know the outcome?

Best,

Claudia Landivar

Community Manager & Editor

[CommunityUser]

@Claudia.Landivar ,

I am following up on a question that nobody properly addressed here - at least it seems like that to me. 

"What if the Application Don't get Sufficent load for last 30 minutes ? Still App agent woud be Going down right ? How to overcome this scenario ?" This was posted two years ago by @Abhilash.C and the point is on "how to overcome this scenario".

In other words, is there a way for us to distinguish whether the agent really has problems or has crashed, or it's just that there is no app load during some period of time? Correct me if I am wrong, but if there is no app data the agent availability percentage will go down for that period of time?

[Claudia_Landiva]

Hi, @Anonymous - I'm working on connecting you with the right PM to answer your question. In the meantime, I have noted it on my ticket to track revising this article.

Claudia Landivar

Community Manager & Editor

[Shubhankar_Yash]

Hi @Anonymous 

Even during no app load, the agents would be sending data to the Controller. As a result, availability wouldn't go down in this scenario. Following are the cases when availability goes down:

1. Agent goes down
2. Agent is unable to connect with the controller (bad config, VPN issue, etc)
3. Server on which the agent is deployed itself goes down

I hope this answers your question.

Regards

Shubhankar

[CommunityUser]

Thanks a lot. That does it.