Edwin AI User Process Flow

Scenario: Checking the health of your systems and managing issues

The following describes a suggested routine for checking system health, and monitoring, investigating, and acting on critical issues.

1. Get an overview
   • Log in to Edwin AI and open the Insights Dashboard to get an overview.
     
   • Check the key metrics at the top. The differential metrics shows the change in number of events compared to the previous time period. Do you see any alarming changes here?
   • Check the number of insights that have been escalated to incidents in ServiceNow.
     
     
2. Review alerts and insights
   • Browse insights and singleton (uncorrelated) alerts – find out if there are any new alerts or insights. 
   • Filter the dashboard view and narrow down issues for example based on severity, tenant ID, or specific resources.
   • Find out if there are any critical issues that need immediate attention?
     
3. Investigate critical issues
   • Right-click the insight to open Insight Details, explore the information in the left panel. 
   • Look at the summaries, associated Tags, Tenant ID, and causal and impacted resources (CI). Can you identify the resource that seemed to have caused the issue?
   • Click the ML Results tab to see what the insight correlations were based on.
4. Explore the issue timeline
   • Look at the timeline view in the right panel of the Details page.
   • Investigate when the alerts started, the frequency, and impacted resources. 
   • In the panel at the bottom, look at the alerts and events associated with the insight.
     
5. Analyze event details
   • On the Events tab of the Details page, check Description to get some additional infrastructure and event context from the ServiceNow CMDB.
6. Do a root cause analysis for the issue
   • Based on the information you got from insight, alert, and event details, continue with a root cause analysis for the issue.
   • Click CI to further investigate the state of the resource in your LM Portal.
     
   • Click Name to investigate the datasource related to the issue in your LM portal.
     
   • Click Source ID to further investigate associated alert information in your LM portal.
     
   • To get more context and to understand the recurrence pattern for an issue, open the Explore page in Edwin AI. 
   • Look at the raw data stream from the data ingested by LM to see if you can identify patterns or anomalies over a specific time range for the issue you are investigating.
     
7. Decide on next steps
   • When you have identified the root cause, use the data you collected to decide on the next steps, for example to escalate the insight and create an incident.
   • Right-click on an insight to manually create a ServiceNow incident. This can also be done automatically depending on your setup of Edwin AI.
     
8. Keep monitoring
   • Continue to monitor the Insight dashboard for changes in system metrics and new alerts.
   • You can set up custom dashboard filters for example to monitor any changes that might indicate potential issues.