How to Check AVEVA PI System Health Across PI Points, AF, Analyses, and PI Vision

AVEVA PI System environments often grow over many years. New PI Points are added, Asset Framework structures evolve, AF Analyses change, and PI Vision displays are copied, edited, and reused by different teams. Over time, it becomes harder for a PI administrator to answer a simple question: is the system actually healthy?

A useful PI System health check should go beyond checking whether services are running. It should help you understand data quality, object health, display reliability, dependency impact, and recent changes across the full PI landscape.

What "PI System Health" Actually Means

A PI System can be online and still have health problems that affect operations. A display may load but show stale data. An AF Attribute may exist but point to a missing or unreliable source. An AF Analysis may be enabled but failing, stopped, or creating unnecessary load.

A practical PI System health check should cover five areas:

1. Start With PI Point Health

PI Points are often the first place to look because they are the foundation for many AF Attributes, Analyses, notifications, reports, and PI Vision displays. If a tag is stale, bad, missing, or outside expected limits, that problem can appear in many places.

When reviewing PI Point health, check for:

• Stale latest values
• Bad latest value quality
• Missing or unavailable PI Points
• Out-of-range values
• Unexpected "No Data" or "Pt Created" states
• Metadata issues such as naming, point source, descriptor, or instrument tag problems

2. Review Asset Framework Health

Asset Framework adds context to PI data, but it also introduces another layer of references, templates, calculations, and inherited structures. A PI System health check should inspect AF Attributes, AF Elements, AF Templates, and AF Analyses together.

AF Attributes

AF Attributes can become unhealthy when their data references are broken, unresolved, stale, or linked to PI Points with poor data quality. It is important to know not only that an attribute is unhealthy, but also which element, template, display, or calculation depends on it.

AF Analyses

AF Analyses should be reviewed for status, input quality, output quality, and service load. A stopped or failing analysis can silently affect downstream displays and calculations. High-load analyses may also deserve attention during performance reviews or cleanup work.

3. Check PI Vision Display Health

PI Vision is where many users experience PI System issues first. A display may look fine at a glance but still contain broken data sources, stale values, missing references, or objects that nobody owns anymore.

For PI Vision displays, check:

• Displays with broken or unresolved data sources
• Displays that depend on unhealthy PI Points or AF Attributes
• Unused or abandoned displays
• Display ownership and last access information
• Datasource counts and missing datasource counts

4. Trace Dependencies Before Making Changes

A health issue is only part of the story. PI administrators also need to understand impact. If a PI Point is stale, where is it used? Which AF Attributes depend on it? Which AF Analyses use it as an input? Which PI Vision displays will be affected if it is changed, renamed, deleted, or fixed?

Dependency mapping is especially useful during cleanup, migration planning, troubleshooting, and template reviews. It helps reduce the risk of changing the wrong object or fixing only the visible symptom instead of the root cause.

5. Look At What Changed Between Scans

Many health problems start after a change: a renamed PI Point, a moved AF Element, a changed data reference, an updated AF Analysis, or a deleted object. Change tracking helps connect new issues to recent activity.

Useful change events include:

• Created objects
• Renamed or moved objects
• Updated metadata or configuration
• Deleted or missing objects
• Excluded objects

Manual PI System Health Check Checklist

If you are performing a manual review, this checklist is a good starting point:

When Manual Checks Become Too Slow

Manual health checks can work for small environments or isolated troubleshooting tasks. But in larger PI System landscapes, the amount of inventory and dependency information can quickly become too large to review by hand.

That is where a scan-driven approach helps. PI Nexus+ scans PI Data Archive, Asset Framework, AF Analyses, and PI Vision to build a searchable view of health, dependencies, issues, and changes between scans.

Note: AVEVA, PI System, PI Data Archive, Asset Framework, and PI Vision are trademarks or product names of their respective owners. PI Nexus+ is developed by Software Athlete.