A Practical Guide To Help With Resolving Future Similar Issues
Problem:
NRIO network on a JACE was showing “ghost” modules from a previous site station copy,(light grey in NRIO Device Manager). These stale devices caused addressing conflicts, discovery kept bringing them back, and IO points on the live (dark grey) modules were not returning values even though sensors were wired.
Final resolution (including what the client did):
Following guidance, the client manually deleted the old NRIO devices from the station database to break the association with the ghost modules. They then re‑added the physical NRIO modules by dragging them from the local station into the NRIO driver on the JACE, ran a new discovery (which no longer returned the ghost devices), and successfully matched the live modules. As a consequence, they needed to relink the IO points that had been disconnected by this clean-up, which they accepted as a quick follow‑up task.
What caused the ghost modules to appear?
They appeared because the station database that was copied still contained NRIO devices from a previous job, and these stale records were being used during discovery instead of only the live hardware.
More specifically:
-
The old NRIO modules from a previous site were never fully removed from the station, so their device entries remained in the database.
-
When the client ran a discovery, Niagara matched what it saw on the RS485 bus against that existing database. The old entries showed as light grey “ghost” devices (stored in the station but not physically present).
-
Those stale database entries then interfered with addressing and matching, so discovery kept “remembering” and re‑showing them.
When the client deleted the devices from the database and re‑added the real modules into the NRIO driver, discovery no longer had those stale records to match against, so the ghost devices disappeared.
We also found that this issue had been confirmed and is documented in Tridium's internal development case reference: NCCB-71360
Key Symptoms and Diagnostics
Devices added via the incorrect view (UX/Web) will present several common issues:
-
Devices do not display Installed Version or Sec Version values.
-
Device health shows status: ok, but the health field reads: fail.
-
The station’s Application Director may show related error messages or exceptions.
-
Points may fail to read/write or appear unresponsive.
Devices configured correctly through the Workbench (AX-style) view will display complete version information, respond properly, and pass health checks.
Correct Configuration Process
-
To ensure proper operation, devices should only be discovered and added using the Workbench AX view of the NRIO Device Manager.
-
This method remains fully supported in Niagara 4.14 and avoids the failure scenarios listed above.
UX View Limitations
The web-based (UX) NRIO view in Niagara 4.14 supports limited functionality. While users can still:
-
View existing devices
-
Add or command points
They should not use this interface to add or configure NRIO devices.
Recommendation
For any station using NRIO on a JACE 8000/9000, use only the Workbench NRIO Device Manager view for device discovery and initial configuration. Avoid using the web interface for these tasks until further notice or an official software update resolves this issue.