IMD SolutionComputing LibraryShop main store

Troubleshooting

Replicate to PI needs PI Web API 1.16 or later

Edge computing and IIoT Published 6 min read

Short answer

Replicate to PI from AVEVA Historian 2023 R2 sets no PI Server version requirement. The only stated floor is PI Web API 1.16 or later on the PI side, and the feature exists from Historian 2023 onward. Read the PI Web API version first; if it reads below the floor, upgrade PI Web API, and treat the Data Archive version as a PI Web API compatibility question, not a replication blocker.

On this page
  1. What the requirements actually pin
  2. Why the floor lands on PI Web API, not the Data Archive
  3. Checking a PI Web API 1.20.0.2008 install against the 1.16 floor
  4. Check the PI target before enabling replication
  5. Reading a failed replication
  6. Fixes that look right but miss
  7. Confirm values arrive, then scale
  8. Questions asked

Replicate to PI from AVEVA Historian 2023 R2 does not pin a PI Server version. The only stated version requirement sits on the PI side: PI Web API 1.16 or later. That makes the compatibility check a PI Web API version check — read the version first, upgrade PI Web API if it reads below the floor, and leave the Data Archive version alone unless the release notes of your PI Web API build say otherwise.

What the requirements actually pin #

Three components sit in the replication chain, and only two carry a stated requirement:

  • Historian 2023 or later. Replicate to PI was introduced in Historian 2023, so 2023 R2 includes it and nothing earlier does. A plant on an older Historian upgrades Historian; no PI-side change adds the feature.
  • PI Web API, at the documented floor. The only version floor in the chain, and the component Historian actually talks to.
  • PI Data Archive, nothing pinned. No PI Server version is stated as required for replication. The effective constraint comes from whichever PI Web API build you deploy: the PI Web API 2023 SP1 release notes, for example, recommend PI Data Archive 2018 SP3 Patch 3 or later and PI AF Server 2018 SP3 Patch 3. Those recommendations attach to that API release, not to Replicate to PI.

Replication keeps PI current from the moment it is enabled. Filling historical gaps that already exist in PI is a separate job, and the guide to backfilling PI from an existing plant historian covers that path.

Why the floor lands on PI Web API, not the Data Archive #

Replication reaches PI through the PI Web API REST interface. Historian speaks HTTPS to PI Web API, and PI Web API speaks to the Data Archive over its own connection. Historian's contract is with the REST surface, so the version floor lands there, and the server behind it is constrained only by what the deployed PI Web API build supports.

The practical consequence decides the order of work: an older Data Archive does not by itself block replication, while an older PI Web API install does. PI Web API is an optional add-on that often stays at the build it shipped with while the Data Archive moves through regular PI upgrades — exactly the situation the version check exists to catch.

Checking a PI Web API 1.20.0.2008 install against the 1.16 floor #

Inputs:

  • Required floor: PI Web API 1.16 or later — the stated Replicate to PI requirement in AVEVA's Historian documentation.
  • Installed version: 1.20.0.2008 — the number PI Web API 2023 SP2 reports, read from Apps & features on the PI Web API host.

Rule, in words: compare the first two numbers of the version — the major and the minor — against the floor. The trailing digits are build and patch identifiers and never lower the feature level.

Arithmetic: 1.20 against 1.16. Major: 1 on both sides, no difference. Minor: 20 > 16, so 1.20 ≥ 1.16, with a margin of 1.20 − 1.16 = 0.04.

Result: the install clears the requirement.

Decision: upgrade nothing on the PI side for version reasons — PI Web API passes, and no Data Archive version needs pinning for this feature. The threshold runs in one direction: at 1.16 or above the host passes; a readout below 1.16 fails, in which case PI Web API is upgraded first and its release notes are read for the supported PI Server versions before installing.

Check the PI target before enabling replication #

  • Read the PI Web API version on the PI Web API host, in Apps & features. At or above the floor, move on; below it, upgrade PI Web API before anything else, then re-check.
  • Confirm the PI Web API installation has Open Message Format enabled; replication writes through OMF.
  • From the Historian node, send an HTTPS request to the PI Web API base URL — PowerShell's Invoke-WebRequest does it in one line. A name-resolution or certificate error here stops replication the same way it stops this test, so surface it now.
  • Match the host name: use in the Historian configuration the name that appears in the certificate subject or subject alternative names, not an IP address or short name. If the certificate is self-signed, import it into the trusted store of the account that runs the Historian service.
  • Prepare the PI-side write account: identify or create the account PI Web API writes with, grant it write access on the target PI points and any objects replication creates, and verify that access from the PI side, not from Historian.
  • Enable replication through the Replicate to PI procedure in the Historian 2023 R2 documentation, entering the PI Web API URL and credentials as that documentation specifies, and start with a single tag.

Reading a failed replication #

Read the actual message in the Historian error log first, then sort the fault into one of four classes before changing anything:

What you see Fault class Check
Nothing connects and no PI points appear Binding: URL, DNS, firewall or TLS trust The reachability request from the Historian node
Connection made, credentials rejected Account The PI-side write account and its mapping in PI Web API
Some tags replicate while others fail Tag The failing point's security, name and data type on the Data Archive
Path and credentials fine, feature calls fail Version The installed PI Web API version against the stated floor

Every tag failing identically points at binding or version. Only some failing points at the tags themselves.

Fixes that look right but miss #

  • Upgrading the Data Archive 'to be safe'. No PI Server version is stated as required for this feature; upgrade PI Web API and retest instead.
  • Short-name or IP mismatch. A URL that does not carry the fully qualified name in the certificate fails trust in a way that looks like a connection fault.
  • Firewall rules on the wrong hop. Historian needs only the HTTPS hop to PI Web API; PI Web API makes the Data Archive connection, so that hop opens from the PI Web API host.
  • Trust under the wrong login. A self-signed certificate imported for your own account is invisible to the Historian service account that actually makes the connection.
  • Chasing the feature on an old Historian. Replicate to PI appeared at Historian 2023; on anything older, PI-side work cannot create it.

Confirm values arrive, then scale #

Replicate one tag first and confirm the matching point exists on the Data Archive. Compare the Historian's latest value and timestamp with the PI point's current value and timestamp, then watch several update cycles and confirm new values keep arriving with advancing timestamps.

Add the remaining tags in batches. After each batch, check the Historian error log for per-tag rejections before adding more, and sample a few replicated points to compare value, timestamp and quality against the Historian source.

That closes the version question as well: with PI Web API at or above the stated floor, OMF enabled and the trust and write account verified, replication runs on whatever Data Archive version your PI Web API build supports.

Questions asked

Where do I read the installed PI Web API version number?

On the PI Web API host, open Windows Apps & features — Programs and Features on older Windows builds — and read the version shown against the PI Web API entry. PI Web API 2023 SP2, for example, reports 1.20.0.2008. Compare only the first two numbers against the 1.16 floor for Replicate to PI; the trailing build digits identify the patch level, not the feature level.

Which PI Data Archive does PI Web API 2023 SP1 support?

Its release notes recommend PI Data Archive 2018 SP3 Patch 3, build 3.4.440.477, or later, together with PI AF Server 2018 SP3 Patch 3, build 2.10.9.593. Those are recommendations attached to that PI Web API release, not the replication floor: the Replicate to PI requirement pins only the PI Web API version. Read the release notes of whichever PI Web API build you deploy before you install it.

Does Open Message Format need to be switched on?

Not on a stock system. The PI side must run PI Web API with Open Message Format enabled, and OMF ships enabled by default, so a default installation needs no change. Verify the setting only if a previous administrator disabled it, and re-enable it before you start replication, because the feature writes its data through OMF.

How much data can Replicate to PI move?

AVEVA's published material reports replication of up to 50,000 PI tags per second without performance degradation, and published large-scale validation work describes the same order of scale with no bottlenecks. Treat both as reported figures rather than plant measurements: your ceiling depends on tag count, change rate and network. Bring tags in batches and watch the Historian error log as you scale.

What hardware does the machine connecting to PI need?

The documented minimum for the component that connects to PI is modest: a 1 gigahertz or faster processor, 4 GB of RAM and 32 GB or larger of hard drive space. Any current industrial PC or box computer clears those numbers easily, so version compatibility and certificate trust decide this project, not compute.

Keep reading

  • How-to

    Backfilling PI from an existing plant historian

    Split the historian-to-PI backfill into four independent stages — extract, intermediate store, external compression, destination load — and qualify throughput with a measured pilot. Rows divided by elapsed seconds, not tag count, sets the schedule: 12,000 tags from 2020 at 500 rows/s is about 203 days. Check the native Historian 2023 replication route before writing pipeline code.

    Edge computing and IIoT

  • Troubleshooting

    Citect SCADA connector to Historian shows disconnected

    A "Disconnected" destination status means the Citect SCADA Connector has no session to Historian and is buffering samples to local disk. SMC tag-discovery and write entries prove the connector's stages, not that values were archived. Stop the connector, verify the destination instance, services and write authorization, reconnect once, and close only when a query returns a sample and the buffer drains.

    Edge computing and IIoT

  • Explainer

    Do you need a historian or just an OPC tag server?

    No. An OPC tag server (OPC DA) supplies the current value, quality and timestamp of each item and stores nothing, so monthly figures need a historian or a SQL logger collecting from that same server. Keep a reporting client like ActiveFactory only if you also run the historian it expects.

    Edge computing and IIoT