How Do You Know When a ServiceNow Upgrade Is REALLY Complete?

[Jeff Boltz1]

Hi Community,

I'm looking for guidance on how other ServiceNow customers determine when an instance upgrade is fully complete and ready for post-upgrade operational validation.

During a recent Australia Patch 5 upgrade, we observed an interesting discrepancy between the Upgrade Monitor and the distributed application node upgrade process.

What We Observed

• Upgrade Monitor reported "Instance upgrade completed successfully."
• The Upgrade Log recorded "Notifying HI that upgrade has completed."
• However, distributed application node upgrades continued to be initiated after that completion notification.
• The Node Upgrades dashboard continued showing nodes transitioning between Running, Failed, and Successful.
• Some nodes initially reported as Failed subsequently transitioned to Successful.
• Approximately 43 minutes after the initial completion notification, the final observed node-chain progression occurred.
• Subsequently, all expected application nodes displayed Successful status.

The upgrade was configured for serial distribution, and the overall process ultimately completed successfully.

 

Why This Matters

Our operational procedures require post-upgrade validation activities, including:

• MID Server availability, connectivity, and version alignment
• Platform health and system log analysis
• Integration connectivity and ECC queue processing
• Monitoring dashboards and operational telemetry
• Automated health checks

The challenge is knowing when these activities should begin.

If validation begins immediately after Upgrade Monitor reports successful completion, application nodes may still be upgrading or restarting.

This can introduce:

• False-positive alerts during expected upgrade transitions
• Misleading health-check results
• Unnecessary troubleshooting and escalation
• Inconsistent operational procedures
• Difficulty reliably triggering automated post-upgrade validation

Conversely, waiting for every node to display Successful may not always be appropriate if some nodes are intentionally offline.

We're working toward an automated, monitoring-by-exception operating model and would like to establish a reliable transition between upgrade execution and post-upgrade operational validation.

Questions for the Community

1. Authoritative Completion Signal

What do you use as the definitive indicator that a ServiceNow instance upgrade has fully completed, including distributed application node upgrades?

2. Expected Upgrade Behavior

Is it expected that distributed application node upgrades continue after Upgrade Monitor reports successful completion?

3. Programmatic Detection

Is there a supported event, API, table, or log entry that confirms all expected application node upgrades have finished?

4. Operational Best Practices

How do you determine when to initiate post-upgrade validation, particularly if you've automated your maintenance procedures?

5. Transient Node Statuses

Are temporary Failed or Down statuses expected during distributed upgrades, and how do you distinguish them from genuine upgrade failures?

 

What We're Trying to Achieve

Ultimately, we're looking for a reliable operational handoff point:

When has ServiceNow finished its upgrade activities so that we can safely and reliably begin ours?

I'd appreciate hearing how other customers approach this, particularly anyone who has implemented automated post-upgrade validation or monitoring-by-exception.

Thanks in advance!