IMD SolutionComputing LibraryShop main store

Troubleshooting

Why a migrated historian shows no PLC connection options

Edge computing and IIoT Published 8 min read

Short answer

A migrated historian shows no PLC connection options because the Allen-Bradley communication driver package was never installed on the new server; the OPC entries in the tree are not that driver. Install the ABCIP package that matches your historian release, confirm the OI Server accepts your license format (.xml or .loc, not .lic), and treat a closed DAServer branch that keeps collecting as an interface fault a restart clears.

On this page
  1. Why does a new historian show OPC servers but no PLC connection options?
  2. How do you install the missing Allen-Bradley driver on a new historian?
  3. Which license format does an OI Server need after a migration?
  4. Why does the DAServer branch stay closed while the PLC keeps updating?
  5. How do you rebuild the PLC connection inside the ABCIP driver?
  6. How do you prove the new connection before moving the tags?
  7. When is a bad value a tag problem instead of a driver problem?
  8. Questions asked

The PLC connection options are missing from the new historian because its Allen-Bradley driver package was never installed: a fresh install lists the OPC interfaces it includes and nothing else. Install the ABCIP driver package that matches your historian release, restart when the installer asks, and the Allen-Bradley branch appears in the configuration console. Check the OI Server's license format while you are at it — one that requires .xml or .loc licensing will not obtain its license from a .lic file. The old server's closed DAServer branch is a different fault: with acquisition still live, it is a stalled management interface, and a restart in a maintenance window clears it. This article separates the two faults, shows the arithmetic that proves whether acquisition is still running, and walks the rebuild up to the first migrated tag.

Why does a new historian show OPC servers but no PLC connection options? #

The tree lists only what is actually installed: a fresh historian includes the OPC interfaces but no controller drivers, so the Allen-Bradley options appear only after the ABCIP package from the Communication Drivers Pack is installed on that server.

OPC entries are interfaces to servers that exist on the machine; the Allen-Bradley controller support is a native driver, a different component delivered by the Communication Drivers Pack. Seeing OPC servers therefore proves nothing about controller support, and selecting an OPC entry does not create an Allen-Bradley connection behind it.

Sort your symptoms before changing anything:

  • Configuration console shows OPC servers but never an Allen-Bradley branch: the driver package is not installed — install the Communication Drivers Pack that matches the historian release.
  • DAServer branch on the old server will not open while PLC values keep updating: management-interface fault — leave the running tags alone and restart in a maintenance window.
  • Driver installed and connection configured, but one item reads wrong or stale: item address, route or tag binding — work the diagnostics before reinstalling anything.

Decision: if the new server has never shown the Allen-Bradley options, the driver install comes before any license or network work.

How do you install the missing Allen-Bradley driver on a new historian? #

Install the Communication Drivers Pack that matches your installed historian release, and restart when the installer asks: AVEVA's driver-pack documentation states a reboot may be required to complete the installation, and the Allen Bradley – ABCIP Driver entry appears in the console only after that completes.

  1. Record the exact historian release and Windows build on the new server.
  2. Obtain the Communication Drivers Pack for that release through the vendor or an authorized distributor — it is a separate install, which is why a default installation leaves it out.
  3. Run the pack's installer and let it add its components, including the ABCIP driver.
  4. Restart the machine if the installer or its documentation calls for it.
  5. Open the configuration console and confirm the Allen Bradley – ABCIP Driver entry now exists before building anything against it.

Decision: the pack plus the reboot is the whole fix for a tree that never showed the options; if the options used to show, skip to the DAServer fault below.

Which license format does an OI Server need after a migration? #

Match the license to the server release, not to the old machine: an OI Server that requires an .xml or .loc license will not take a .lic file, so software that installed cleanly can still run without authorization and fail in ways that look like a network problem.

The format requirement belongs to the release, so confirm what the installed release accepts before importing anything. Pair the driver and server versions from the vendor's compatibility information at the same time — the same per-release check that picks the driver pack.

Decision: driver installed, connection configured, connection will not start — check the license format and entitlement next, not the network.

Why does the DAServer branch stay closed while the PLC keeps updating? #

The configuration tree is a management view, not the data path: a running DAServer keeps collecting on the configuration it loaded at startup even when its branch refuses to open, so live values with a dead tree describe an interface fault, not an acquisition failure.

Confirm which fault you have before touching anything. Read the newest timestamp the historian holds for a tag the PLC updates on a fixed interval, and compare it with the current time.

Worked example, illustrative numbers: a counter tag updates every 2 s, and server and controller clocks agree. The controller reads 14:32:05; the newest stored timestamp for the tag reads 14:26:05. Staleness = 14:32:05 - 14:26:05 = 6 min = 360 s. At one update every 2 s, 360 s ÷ 2 s = 180 consecutive missed updates. No console glitch explains 180 silent scans: acquisition has stopped, and the closed branch is a second problem, not evidence that data still flows. Read the cutoff off the tag, not off this example — a few update intervals of staleness means acquisition is live; hundreds of intervals means it is not.

If staleness sits within a few update intervals, the branch is an interface fault. Restart the communication server service, or the machine, in a maintenance window: acquisition stops until the server reloads, so never restart it blind while it feeds operator screens or an alarmed process. A console object that will not open is an interface-layer fault, and the same separation applies when a historian device shortcut that will not appear is the only symptom.

Decision: live values with a closed branch — restart in a maintenance window; dead values — treat it as an acquisition stop and start at the driver and license, not the console.

How do you rebuild the PLC connection inside the ABCIP driver? #

Rebuild the connection in the driver layer before you touch a single tag: AVEVA's ABCIP documentation has you configure port objects for the Allen Bradley – ABCIP Driver and then add and configure an AO, and the PLC address and any controller route belong at that level.

  1. Record the old connection's address and route from the old console, if it opens again, or from your controller and network records.
  2. Create the port objects for the Allen Bradley – ABCIP Driver in the new console.
  3. Add and configure an AO for the controller with the same IP address and route the old connection used: the IP identifies the network endpoint, the route identifies the controller behind it.
  4. Activate the configuration using the server's documented activation step.
  5. Open the driver's diagnostics and read one known item; confirm value and quality before anyone mentions tags.

The OPC entries can serve as the acquisition path only where a licensed OPC server is already configured and serving those items; otherwise the rebuilt ABCIP connection is the path.

Decision: old tags came through the ABCIP driver — rebuild in ABCIP; a licensed OPC server already serves the items — use it and skip the driver work.

How do you prove the new connection before moving the tags? #

Migrate one tag first, not the database: bind a single known item, confirm the historian stores the same value, quality and timestamp the driver reports, then query the stored history back out — a bulk move turns one bad binding into a hunt across hundreds of tags.

Bind the test tag with the server reference and item address the driver publishes, compare the three fields against the driver's diagnostic read, then query the tag's stored history from the new store.

A clean driver read does not prove the value reached the store: historian pipelines can drop or clear stored entries when their source changes — the failure mode that store-and-forward entries clearing when their source changes documents. Check the store, not just the live read.

Configuration does not carry the archives across, and moving the old history is a separate job — backfilling PI from an existing plant historian covers that path once the live connection is proven.

Decision: driver value, stored value and queried history all agree with good quality — migrate the remaining tags in batches; any disagreement — stop at the failing layer before another tag moves.

When is a bad value a tag problem instead of a driver problem? #

Compare both ends of the chain and the split names the layer: a correct driver value with good quality beside a stale or missing historian copy is a tag binding fault, while both ends wrong points upstream to the connection or the controller.

  • Driver correct with good quality, historian stale or missing: fix the tag's source reference or item binding — the item address must be written the way the chosen acquisition path expects, and a native driver reference is not the same string as an OPC item reference.
  • Both ends wrong: work backward through the server runtime and license, the configured address and route, network reachability, then the controller's communication state. A successful ping proves only that an endpoint answered, not that the driver can read your item.
  • Value arrives but reads wrong: that is scaling, not connectivity — historian tag scaling for an unscaled PLC integer covers that fix.

Once tags store correctly, a trend client that cannot connect after the migration is a separate client-server fault, not an acquisition one — historian trend client connection failures have their own troubleshooting path.

Decision: driver right and historian wrong — fix the tag binding; both wrong — start at license and address, not at the tag database.

Questions asked

How do I know which driver release matches my historian version?

Driver releases are published against specific software versions, so compatibility is checked per release, not in general. AVEVA's Plant SCADA documentation also notes that a required driver may be missing from an installer when it is new, which is why the pack is a separate download for your installed release. Match the pack to the historian release and operating system you recorded, and take the version pairing from the vendor's compatibility information for that release.

How can I confirm a license actually installed on the new server?

AVEVA's licensing documentation notes that the check and install of the license file is written to a console.log file stored in the default folder under C:\Program Files (x86)\AVEVA\. Read that log after importing the license: a clean entry means the file was accepted, and a failure line tells you whether the format or the entitlement was the problem before you spend time on the connection itself.

The branch came back after a restart — can I trust it, or will it disappear again?

If the branch disappeared while values kept updating, the restart addressed an interface fault and the data path was never broken, so a branch that stays open afterwards is a normal working console. If it closes again, note what ran just before it failed — an activation, an update, a network change — and take that pattern to the vendor's support, because a recurring interface fault deserves a cause, not another restart.

Does installing the new historian move the old trend data across?

No. The old store keeps its archives until you decommission it and the new store starts empty; installing drivers and rebuilding connections moves configuration work, not history. If reports need continuous history across the changeover, plan a backfill from the old historian into the new one after the live connection is proven, and verify the backfilled ranges against the old store before retiring it.

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

  • How-to

    Deploying Edge Data Store in Docker on a controller PC

    Run Edge Data Store as a Docker container on the controller-class edge PC and let its OPC UA adapter ingest the local server over localhost. Define egress targets, buffering and forwarding schedules in the JSON configuration, size the buffer from tag count, scan rate and outage target—50 tags at 1 event/s across an 8-hour outage is about 288 MB—and sign off only after a live pulled-cable test.

    Edge computing and IIoT

  • Troubleshooting

    Why Power BI cannot connect to FactoryTalk Historian

    Power BI has no native FactoryTalk Historian connector because the PI-based archive is a proprietary time-series store, not SQL tables. Reach it through the licensed 64-bit OLE DB or ODBC client provider installed on the Power BI machine, aggregate on the server before import, or stage the data through an Excel PI DataLink workbook when no provider is licensed.

    Edge computing and IIoT