Here’s a sample workflow to demonstrate the use of the Orchestration app to automate remediation of GCP instance.


Executing self-healing workflows based on alerts coming from the monitoring tool(New Relic), to fix the problems through resetting the instance to its original state and making the GCP instance available. 


Assumptions

  • New Relic alert profile must be integrated with Freshservice 

  • Alert rules to tag the relevant instances. For this use case, we’re assigning the tags

    • P0_alert 

    • instance-1 

    • us-central1-a 

We’ll use these tags within the app action node to reset the correct instance.


Note: To integrate New Relic APM with Freshservice Alert Management, follow the documentation provided. 


Workflow configuration

  1. Create the Event when an Incident is raised. Followed by a condition node to check whether the incident is p0 priority. 

  2. You can then pull in the App Node which will have the following configurations:
    App Name: GCP - Orch

App Action: Reset Instance

Note while configuring inputs:

To specify the name and zone, we have to use liquid templates to map ticket tags and to fetch the appropriate value.


For example,

Placeholder for the name will be, 

 {% assign value = ticket.tags | split: ", " %}{{value[1]}}


Placeholder for the zone will be, 

{% assign value = ticket.tags | split: ", " %}{{value[2]}}


  1. To validate whether the instance reset is successful or not a condition node is added with a check for status code as 200.

  2. Finally, add a public note with a confirmation message


GCP Activity Console