aveva pi

Latest version isn't always the greatest

March 23, 20262 min read

Why Your OPC UA Adapter Upgrade Broke — And How to Fix It

Upgrading to the latest version should make your system better.

But in industrial environments — especially with OPC UA — that assumption can cost you hours.

I recently upgraded the AVEVA Adapter for OPC UA from v1.4.1.32 to v1.6.0.296.

Everything looked fine.

Until the tags stopped updating.

No obvious errors. No clear logs. Just silent failure.


The Real Problem (That Isn’t in the Release Notes)

After hours of troubleshooting, the root cause surfaced:

"Mismatch of Endpoints returned by GetEndpoints and CreateSession"

The newer adapter version (1.6) enforces stricter endpoint validation.

If the OPC UA server returns an endpoint URL that doesn’t EXACTLY match what the adapter connects to — the session is rejected.

Even small differences break it:

  • Hostname vs IP address

  • Case sensitivity

  • Load balancer altering the endpoint path

Version 1.4 tolerated this.

Version 1.6 does not.


Why This Matters

This isn’t just a bug — it’s a protocol compliance issue.

The OPC UA specification expects consistency between:

  • Endpoints returned by GetEndpoints

  • Endpoints used during CreateSession

Some servers are loose with this.

Newer clients are not.

That mismatch = broken data flow.


What Actually Works

Short-Term Fix

Downgrade to AVEVA Adapter for OPC UA v1.5.

This version is less strict and restores functionality.


Long-Term Fix

Fix the root cause at the OPC UA server level.

Your server must return consistent EndpointDescriptions across:

  • Discovery (GetEndpoints)

  • Session creation (CreateSession)

If not:
→ You will keep breaking with stricter clients.


Proof This Isn’t Edge Case

This issue exists in a buried support article.

Not in release notes.

Which means:

  • Most teams will discover it the hard way

  • During production upgrades


The Real Lesson (Most Teams Miss This)

Don’t assume backward compatibility in industrial protocols.

Especially when vendors improve standards compliance.

Before any upgrade:

  • Test in a dev environment

  • Validate handshake behavior

  • Compare endpoint responses

  • Simulate real connection paths


Want the Exact Validation Checklist?

I put together a simple checklist to validate OPC UA adapter upgrades before deployment.

It covers:

  • Endpoint validation steps

  • Common mismatch scenarios

  • Pre-upgrade testing sequence

[Download the checklist here]


Final Thought

Hours burned — not because of complexity, but because of hidden assumptions.

If you're working with OPC UA integrations, this is the level of detail that separates stable systems from unpredictable ones.

And if you're scaling industrial data infrastructure, these edge cases aren’t rare — they’re inevitable.

The only question is whether you catch them early — or in production.

https://www.linkedin.com/posts/rickykwokshingsun_pisystem-aveva-piadapter-activity-7439683231265939456-9LYh?utm_source=share&utm_medium=member_desktop&rcm=ACoAAA-EXgYBnbnWXee4O7r8oubyI7D5gid3Op8

Ricky Sun is the founder of SunLead Technologies, an industrial data infrastructure engineering firm based in Calgary, Canada. He specializes in modernizing operational data systems for energy companies, with deep expertise in industrial historian platforms including AVEVA PI System, Canary Historian, and modern time-series architectures.
Ricky’s work focuses on PI System migrations, Asset Framework (AF) restructuring, and historian modernization executed inside live production environments. He works closely with system integrators and energy operators to stabilize industrial data infrastructure, improve operational visibility, and prepare operational data for advanced analytics and AI.
Through technical content, industry collaboration, and conferences, Ricky actively contributes to the industrial data and Industry 4.0 community, helping organizations build reliable data foundations for digital operations.

Ricky Sun

Ricky Sun is the founder of SunLead Technologies, an industrial data infrastructure engineering firm based in Calgary, Canada. He specializes in modernizing operational data systems for energy companies, with deep expertise in industrial historian platforms including AVEVA PI System, Canary Historian, and modern time-series architectures. Ricky’s work focuses on PI System migrations, Asset Framework (AF) restructuring, and historian modernization executed inside live production environments. He works closely with system integrators and energy operators to stabilize industrial data infrastructure, improve operational visibility, and prepare operational data for advanced analytics and AI. Through technical content, industry collaboration, and conferences, Ricky actively contributes to the industrial data and Industry 4.0 community, helping organizations build reliable data foundations for digital operations.

LinkedIn logo icon
Youtube logo icon
Back to Blog