compatibility issues between system platform v23 and kepware

Hi,

I have been having lot of problems with a system platform 23 setup which seems to be triggered by suitelink connections, I am using the newest version of kepware as an OPC. It is hard to pinpoint what the reason is but during my research a noticed there is a reported issue from PTC and Aveva regarding problems with 23 and kepware.

here is one example:

Compatibility Concerns: Users have reported difficulties in establishing OPC communication between Kepware servers and System Platform 2023, especially when the servers are on separate machines. Even after configuring DCOM settings and firewall rules, connections may fail. Installing Kepware locally on the same machine as System Platform IDE often resolves these issues, suggesting potential network or configuration challenges in distributed setups.

also here

Integrating Kepware with AVEVA's System Platform can present certain challenges. Notable issues include:

  1. DCOM Hardening Patches: Microsoft's Distributed Component Object Model (DCOM) hardening updates have impacted OPC-DA communications, affecting components like FS Gateway and connections to Kepware. These patches enhance security but may disrupt existing configurations. AVEVA has provided guidance on addressing these changes, recommending registry adjustments and thorough testing in non-production environments.

  2. Compatibility Concerns: There have been reports of compatibility issues between KEPServerEX and AVEVA System Platform 2023. Users have experienced challenges in establishing seamless communication between the two platforms. PTC has acknowledged these concerns and is investigating to provide solutions.

  3. System Failures: Instances where Kepware software ceases to function after a system failure have been documented. Such failures can corrupt the software or its licensing. The recommended solution involves reinstalling the Kepware software and, if necessary, contacting technical support for license reactivation.

My question is:  do we have some others here on this forum that have info or experiences related to something similar?

  • Hi   , I have done the exact above but with a slight change, I moved from OPC DA to OPC UA ( which Kepware and SP 23 support at no additional cost ) and problem solved. The issue you have stated is not Kepware or AVEVA but it's a Microsoft issue no longer supporting OPC DA without compromising on security

  • ok, thank you!

    did you start by using the ddesuitelink object? How did the issues look?

    I am referring to a problem related to the ddsuitelink object:

    • An issue has been observed with multiple versions of KEPServerEX / ThingWorx Kepware Server when interacting with Aveva System Platform 2023 when using the FastDDE/SuiteLink interface
      • The issue is being investigated 
      • At this time Aveva recommends utilizing OPC UA for connecting to the Kepware OPC UA interface
  • There was bad OPC quality data when the OPC in Kepware worked fine, but i tried a separate OPC client and still didnt work, figured this is not an AVEVA issue

  • ok I am having problems which i think has to do with suitelink connections, I am going to test with opc ua instead. 

  •   hi again, I configured one instance of opcclient to connect to kepware, it works fine, but how do I know if i am using opc da or ua, I assumed I should put the path from the OPC UA server (on the screenshot below .) somewhere in the instance settings.

    This works if I put this data in the opcclient instance , (screenshot below:)

    Is this configured as OPC UA connection?

  • Hi   you need to first add OPC UA connection in Operations Management console 

    Please see the link below for detailed instructions

    https://docs.aveva.com/bundle/sp-appserver/page/689813.html

    I would love to help you further but I need time and currently super busy

  • Hi Jakob,
    The Application Server OPC Client object does not support OPC UA (yet) so I would say that in that case you are using OPC DA connecting to your Kepware server.
    To have System Platform connect to a OPC UA Server you need to involve the OI Gateway using a OPC UA Connection node.

    And then use a communication object (preferably the Suitelink object or the new PCS connection protocol (New in 2023).

    OPC UA requires some additional configuration that you will not find in the OPC Object

    And enables new features other than DA (Data Access) such as Methods.

    https://www.youtube.com/watch?v=5Yo7tFE42Jc

    As you and Rainer mention it is advised to move away from OPC DA over network due to the Microsoft DCOM Hardening initiative witch in some cases makes it impossible to communicate with a 3rd party OPC DA Server over network. A workaround could be to install OI Gateway locally on your OPC DA server node and use Suitelink/PCS to manage the node to node communication.

    (PCS removes the need of a Application Server communication object (OPC/Suitelink_DDE), but requires some additional configuration if you have not implemented it before).

  • Hi Richard thank you for your response,

    I am trying to figure out what issues there have been with the suitelink client and kepware this notification is published on their webpage

    • An issue has been observed with multiple versions of KEPServerEX / ThingWorx Kepware Server when interacting with Aveva System Platform 2023 when using the FastDDE/SuiteLink interface
      • The issue is being investigated

    I was hoping some of the good people on this forum have also had issues with the same.

    I am in a nightmare situation with a system for our client that puts items on quarantine randomly as it seems, but I have at least figured out it happens as soon as I deploy the suitelink connections, I can run appengines with deployed instances, when I exclude the Suitelinkconections.

    I then rambled upon this notification from kepware about issues.  Yesterday I was planning on changing to the opcclient and test it that way, but now after explanation from you and Ranier it seems that is also a problem.

    The newest hope is deactivating this option in the kepware settings that I also just noticed, screenshot

  • Outch, yes, I recall you mentioning this issue in another post. Seems to be a rough problem.
    I don't have any quick ideas on this exact issue since I'm not using Kepware in any of our active installations at the moment. But we are planning an upgrade for a customer to 2023 R2 SP1 that includes Kepware so it is of course of interest to understand this issue.

    But I do recall some issues that a college of mine had where he was unable to deploy to the same node as the Omniserver driver, if the driver service was turned off it worked, when started, it crashed the entire server (SMC full of Errors) and we could not get it running until server was rebooted (sometimes several reboots was required).

    The solution was to move the Omniserver to a client node without any Engines deployed.
    Both of these servers (Kepware and Omniserver) has a OPC UA interface exposed, but in Omniserver was not possible to disable the OPC UA part, so perhaps this could be tried, Also, I would try to turn off Suitelink if that is possible too. (to see if it makes a difference).

    Is the Kepware server installed on a node that has Bootstrap and are you deploying any platform to the machine? when you say that you can not deploy Suitelink instances, are those deployed to a stand alone object server, or is it the same node as where Kepware is running?

  • Is the Kepware server installed on a node that has Bootstrap and are you deploying any platform to the machine? when you say that you can not deploy Suitelink instances, are those deployed to a stand alone object server, or is it the same node as where Kepware is running?

    I have tried several different scenarios but now the kepware server is on the same node as the appengine hosting the suitelink connections. 

    I can deploy the suitelink instances, but I get items going on quarantine one by one until everything is in quarantine.

    I disabled the The Dcom config settings on the screenshot from last post, I have had my test connections running for at least 3 hours now which is a record.  So i am hoping that tomorrow when i return to work it is still running. 

    will keep you posted, and thank you!