Remediation Action if Machine Agent is down

[Vishnu Kumar.D R]

Hi All,

We have implemented AppD Server Monitoring with Linux Monitoring Extension in hopes to replace our Nagios Monitoring setup.

Ref: https://docs.appdynamics.com/display/PRO43/Remediation+Actions

Am aware that for a remediation action to run, machine agent should be running.

So in case we have a MachineAgent itself down, how do we enable it back using a script?

Note: Would prefer if no third-party tools are used for this solution.

[Arun.Dasetty]

Hi,

We have feature to monitor other JVM crashes and will report as event (jvm crash guard event) but this reported by machine agent itself, for the scenario you have listed we do not see an option apart from defining HR alert on machine agent availability metric to get UI level alert.

However i do not see any option to enable(restart machine agent) machine agent java process if down as of now, given that we are running only one machine agent java process on that machine.

I have below suggestion: (just basic shell script)- create a bat/shell script to monitor machine agent java process (example: ps -ef | grep machine-agent.jar) and the script will run machine agent startup command (example: java -jar machineagent_dir/machineagent.jar) if the script found in shell logic sh file that machine agent jvm process is down to enable(restart jvm) back agent, see if that will help.

[Vishnu Kumar.D R]

Thanks for the reply.

But that will be the traditional approach, we take care of processes right...

Is there any way we can implement with AppD using a Custom Metric or any other method?

[Arun.Dasetty]

Hi,

I do not see other option as of now for scenario where machine agent itself is down, however can you go through below feature doc https://docs.appdynamics.com/display/PRO43/Add+Tasks+for+Workflow+Steps

[Vishnu Kumar.D R]

Hi,

This sounds good, but ours is a On-Premise deployment.

Can Cloud Auto-Scaling be enabled in our setup?