PI Nexus+ / Reference
PI Adapter Issues and Runtime
This reference lists the PI Adapter statuses, issues and thresholds. The User Guide chapter PI Adapters links here.
Overview
This reference lists the PI Adapter statuses, issues and thresholds. The User Guide chapter PI Adapters links here.
Collector health columns
| Column | Good | Warning | Error | Not Evaluated / Unknown |
|---|---|---|---|---|
| Configuration | The adapter's health elements in AF have the attributes, PI Point references and types PI Nexus+ needs | Not used | A required monitoring attribute is missing, ambiguous or incompatible, or its PI Point reference is malformed or not found | Not Evaluated: the AF structure is not recognized |
| Data Health | Every associated PI Point is healthy | An associated point has a warning | An associated point is bad | No point could be associated |
| Runtime | Device status Good and heartbeat on time | See device statuses below | See device statuses below | Unknown: telemetry unreadable, unbound, bad quality, unrecognized, or monitoring stalled |
A PI Point counts toward Data Health only when its OMF data source matches the Collector's data source exactly. Points claimed by two Collectors are shown as ambiguous and left out. Configuration checks are not tracked issues and cannot be suppressed; they are listed under Monitoring setup on the Collector's Issues tab.
Device statuses
| Device status published by the adapter | Runtime |
|---|---|
| Good | Good |
| Starting, ConnectedNoData, NotConfigured, AttemptingFailover | Warning |
| DeviceInError, Shutdown, Removed | Error |
| Any other value | Unknown |
PI Adapter issues
Only Runtime findings become issues. They appear on the Issues page under PI Adapters and on the Collector's Issues tab. The issue type is PI Adapter Runtime Health; the message names the finding.
| Finding | Issue severity | Raised when |
|---|---|---|
| Device status Warning | Warning | The device status is one of the Warning values |
| Device status Error or Removed | Error | The device status is one of the Error values. A Warning that becomes an Error is the same issue, escalated |
| Heartbeat overdue | Error | The adapter missed its published next-health-message deadline by more than 2 minutes |
| Telemetry bad quality, unreadable, unbound or unrecognized | Warning | The device status or heartbeat could not be used |
| Restart loop (service) | Warning | The adapter service started 3 or more times within 24 hours. A single restart raises nothing. Counted from the start times stored in the PI archive |
| Low storage (service) | Warning below 10 % free, Error below 5 % free | The disk the adapter stores its data on is filling up. The adapter buffers on this disk while PI is unreachable |
Service issues are raised on the adapter service, appear on every Collector of that service, and do not change a Collector's Runtime. They resolve automatically when the condition clears; a failed or out-of-date read neither raises nor resolves them. CPU use is shown on the Service tab for context and never raises an issue. Data Health and Configuration findings do not become adapter issues; the affected PI Points raise their own issues.
A suppressed Collector sends no runtime email. Runtime emails are set under Admin > Notifications > Delivery Rules > PI Adapter runtime and follow the PI Interface rules: they wait for a confirmation window, repeat an open alert only on escalation, and count a stalled Collector as Unknown.
Thresholds and timings
These are fixed in the product, except the poll interval.
| Setting | Value |
|---|---|
| Runtime poll | Interface and adapter runtime poll (see Scan Automation in the Administration Guide), default every minute |
| Monitoring stalled | No successful read for 3 poll intervals and at least 10 minutes; Runtime shows Unknown and the Runtime card counts it as not refreshing |
| Heartbeat grace | 2 minutes after the adapter's published deadline |
| Restart loop | 3 starts within 24 hours |
| Low storage | Warning below 10 %, Error below 5 % free |
