IMDTouchKnowledge BaseShop main store

Troubleshooting

Why FactoryTalk View SE Clients Wireframe While Servers Show Active

HMI and touchscreens Published 7 min read

Short answer

Wireframed FactoryTalk View SE clients with servers showing Active mean live data cannot be acquired from the data providers, and an Active server state never proves that path. Prove the path in a blank local SE project with one FactoryTalk Linx server and one shortcut, then rebuild the converted application in tested increments; the reported recovery came only from a ground-up rebuild with imported graphics and manual shortcuts.

On this page
  1. What the wireframe actually reports
  2. What did not restore data
  3. Isolate in this order
  4. The blank local project test
  5. Rebuild in tested increments
  6. Known software faults worth one check each
  7. The decision, case by case
  8. Questions asked

Wireframed graphics on FactoryTalk View SE clients while every server reports Active means the displays cannot acquire live data from their data providers, and the Active state proves nothing about that path. Treat the wireframe and the frozen tag browser as faults in the application and its references, not in the server installations, and stop repeating reinstall, rollover and Linx replacement: in one reported distributed system, none of those restored data. What worked was isolation first, a blank SE project on one machine hosting every server role, one FactoryTalk Linx server, one shortcut, one known tag, then a ground-up rebuild with graphics imported and shortcuts configured manually, in tested increments.

What the wireframe actually reports #

Rockwell troubleshooting documentation for View SE defines the symbol: wire-framed objects on a display mean the live data items could not be acquired from the data providers. In a distributed SE system that chain has several links, and the wireframe names none of them:

  • the HMI server holding the tag database,
  • the data server configuration feeding it,
  • FactoryTalk Linx and the shortcut that resolves to the controller,
  • the OPC server when one sits in the path,
  • the network route to the controller,
  • the client path back to those servers.

An Active server state in FactoryTalk View Studio is a server status. It is not a tag-resolution test, not a shortcut test and not a value-delivery test, which is why every server can report Active while every client wireframes. The reported clients also raised no lost-connection errors; when comms genuinely drop, clients and panels flag it rather than falling silent, and question marks on live values from PLC comms timeouts are the classic symbol. No error plus wireframe points at the reference and data-provider layer, not at the network drop you would first suspect.

What did not restore data #

In the reported distributed version 9 installation, each of these was tried and none changed the symptom:

  • Reinstalling FactoryTalk View version 9 and reapplying its patches.
  • Upgrading FactoryTalk Services to version 6.5 across the servers.
  • Forcing rollover to the redundant backup, which wireframed identically.
  • Deleting the FactoryTalk Linx server, building a new one and configuring shortcuts manually.
  • Rebuilding the environment on Windows Server 2022 with FactoryTalk View V15 and converting the application, which still wireframed and whose tag browser still froze.

None of those produced a new diagnostic result, so repeating them spends outage time without moving the fault. The identical failure on primary and backup is itself a result: both nodes share the project, the directory data and the communication configuration, so the fault lives in shared content, not in one machine installation.

Isolate in this order #

Change one variable at a time and write each result down before the next step.

  1. Ping the controller from the development machine. That proves IP reachability from that one machine; it does not test directory access, Linx browsing, shortcut resolution or delivery to a client.
  2. Open the FactoryTalk Linx diagnostics and attempt a browse or a live tag while watching the messages; confirm the intended controller is visible and the shortcut points at it. A tag browser empty after shortcut setup has its own checklist, and a browser that freezes instead of opening is a different observation, so record which one you have.
  3. Run the ME application that uses the same communication path and the same PLC shortcut. In the reported case it browsed and displayed normally, which clears the controller route and the shortcut. It does not test the SE project or the SE client runtime; when an MER download shows no live data, the fault sits in that runtime communications configuration, which is why the ME result narrows the SE side but never clears it.
  4. Build the blank local project test that follows.
Observation What it proves Next check
Ping replies from the development PC IP reachability from that machine FactoryTalk Linx diagnostics
ME application works on the same path and shortcut Controller route and shortcut are good SE project and SE client path
Blank SE project shows a changing tag Machine, Linx, shortcut and client runtime are good Converted application content
Converted application wireframes on that same machine Fault is in the project or its references Incremental rebuild
Primary and backup fail identically Fault is in content shared by both nodes Stop server-side reinstall work

The blank local project test #

This is the test that separated the layers in the reported recovery. Host all the server roles locally on one development machine, the single-host pattern whose limits are covered for multiple HMI projects on one server host, and keep it off the external directory at first.

  1. On the development machine, create a new SE application and manually add the server roles, hosted locally.
  2. Configure one FactoryTalk Linx server and one shortcut to one controller.
  3. Browse one known tag, place it on a display and run a test client on the same machine.
  4. Pass only when the displayed value changes while the tag changes in the controller; a browser that lists tags is not runtime proof.
  5. Then point the environment at the FactoryTalk directory and repeat the browse and runtime checks before adding anything else.

Why one machine convicts or clears the project, using the reported installation counts: the failing system ran 4 server machines, a primary hosting the HMI, data, alarm and OPC server roles, plus a dedicated directory server, an alarm and event log server and a historian server, serving up to 12 clients across a managed OT network. That is 4 + 12 = 16 machines, or 17 elements in the failing chain once the network itself is counted. The blank local test exercises 2 elements: one machine and one controller path. It removes 17 - 2 = 15 of the 17, about 88 percent of the fault surface, in a single step. The counts belong to that installation; the method is what carries over.

Keep the emergency local SE stations in service until the rebuilt distributed system proves itself on live displays: wireframed clients leave operators with no live values while the process keeps running.

Rebuild in tested increments #

The reported recovery came only from a ground-up rebuild, so structure the rebuild as the diagnostic:

  1. Keep the blank working project as the control, and never copy the failing project into it as one untested change.
  2. Import one small group of graphics, then play them live in Studio and in the test client.
  3. Recreate the shortcuts manually from the plant verified shortcut list, and retest browse and runtime after each batch.
  4. Add the remaining displays and configuration in batches; when the tag browser freezes or a display wireframes again, remove the last batch and test its contents alone, because that batch holds the damaged content.
  5. After the local build works end to end, connect to the FactoryTalk directory, re-add the server roles there and repeat the runtime checks before pointing production clients at it.

In the reported rebuild all screens were imported this way, FactoryTalk Linx was reconfigured manually from the plant shortcut list, the servers joined the directory, the PLC shortcuts functioned and client testing followed. One DCOM configuration issue appeared when a Kepware OPC server joined the path; once it was resolved, the graphics worked. Expect DCOM whenever a third-party OPC server enters a path that previously worked without one.

Known software faults worth one check each #

Clear these documented, project-independent causes before blaming the project:

  • A race condition in the FactoryTalk View tag browsing engine when communicating with RSLinx Classic is a documented cause of a tag browser that opens and then stops responding, addressed by a vendor patch. If the browser freezes in every project, including a blank one, patch the installation before touching the project.
  • A documented fault, fixed by a patch, stops HMI tags from updating after several hours or days of runtime, so clients keep showing values that no longer change. Relevant when values freeze rather than wireframe.
  • A corrupted HMI server tag cache is a documented cause of ghosted, inactive values; the documented fix is to stop FactoryTalk View SE and delete the tag cache so it rebuilds.

The decision, case by case #

  • Blank project shows a moving tag while the converted application wireframes on the same machine and the same shortcut: the converted application is the fault. Stop reinstalling software and start the incremental rebuild.
  • Blank and converted applications both fail on one machine: the project is cleared for now. Fix that machine Linx configuration, directory connection and client path, and check the route for duplicate addresses.
  • The blank project works locally but fails once the directory is connected: the fault is directory-side, in membership, server references or access. Correct it there before rebuilding anything else.
  • Primary and backup fail identically: the fault is in shared content. A rollover cannot clear it and a single-node reinstall cannot fix it; the incremental rebuild above is the way out.

Questions asked

A ping to the PLC returns in under 1 ms. Is the network proven?

It proves IP reachability from that machine and nothing else, not directory access, Linx browsing, shortcut resolution or tag delivery. Compare the reply time with a ping from a machine that browses successfully. A reply dramatically faster than the route warrants can mean a different device is answering, such as a duplicate address, so check the ARP entry for that IP before trusting the ping.

The site had a power outage before the failure. Did it corrupt the project?

Treat the outage as a timeline marker, not a diagnosis. Corruption is possible if files were being written when power dropped, but the reported virtual servers were believed to have stayed online, and the isolation test, not the outage story, is what identified the project. Compare project, directory and communication changes around the outage time, then run the blank-project control before concluding anything.

Why would the tag browser freeze only in the converted project?

The browsing engine walks the project references, and documented triggers for data server errors include specific data server configuration settings and tag syntax. A blank project that opens normally on the same installation is the control: a freeze that follows the project rather than the installation is what convicted the converted application in the reported case.

Can the converted application be repaired in place instead of rebuilt?

In the reported case only the ground-up rebuild restored graphics and working shortcuts; re-conversion, reinstalls, rollover and a new Linx server all failed. The batch method turns the rebuild into a search: the batch that breaks the browser or wireframes a display contains the damage, and that content is what you fix or replace rather than re-converting the whole application again.

Keep reading

  • How-to

    Sending the Active HMI User to the PLC With Login Macros

    FactoryTalk View has no system tag that copies the logged-in user or group into the controller. Create one login macro per Runtime Security group, each writing its group name to the same PLC STRING tag, and a logout macro that clears it. The PLC then reads a role, not a password.

    HMI and touchscreens

  • Troubleshooting

    Why an MER Download Shows No Live Data on the HMI

    The terminal runs the new MER against the communication configuration already stored on it, because the transfer skipped Replace Communications. Every tag returns Bad quality, so screens draw but show no live data. Re-download with Replace Communications checked, let the terminal reboot once, then prove the path with a heartbeat count.

    HMI and touchscreens

  • How-to

    Relaying serial PLC data through a C-more HMI to SCADA

    The C-more EA9 is a communications client: it polls the PLCs it is configured for and exposes nothing for Ignition or KepServer to browse. Make the panel the relay instead — add a Modbus TCP server as a second PLC device and copy FX3 tags into it with Tag Copy events triggered from the system clock tags, since scheduled copies run at best once per minute.

    HMI and touchscreens