Our customer is developing a program based on the EAL4C sample project to control three ctrlX DRIVEs. They are currently experiencing an issue where the connection between the drives and the PC intermittently drops, causing the servos to switch off.

The issue occurs on all three drives, but not simultaneously.

When it happens, ping still works, but Drive Engineering cannot connect and shows a (no known host?) message. After a short while, the connection becomes available again.

The PC has three physical LAN ports, so each drive is connected to its own port.

What could be causing this?

I'm setting up the additional subnet on the IXON site. I am getting an error relating to the Gateway address being the same as the LAN IP Address. In the past i have not had any issues with the configuration.

History

24 Sept 2026 Initial release

Content

This article describes how to subscribe to ctrlX Data Layer nodes in an HTML document or in HMI systems (HELIO, WebIQ) with JavaScript support. General descriptions and further information can be found in our online documentation.

Requirements

• ctrlX OS device, ctrlX CORE or ctrlX COREvirtual

• HTML and JavaScript know how

For an easy startup, download subscription-demo.zip (see chapter Downloads) it contains:

• subscription-demo.htm - A simple HTML page that subscribes to two data layer items

• cxApi.js – Contains two JS classes

  • cxApi - Manages authentication

  • cxSubscription - Manages the subscription

Quick Startup in 3 Minutes

1. Add a new data layer item.
   Search for text #1# and uncomment the text shown below.

   <!-- #1# -->
   <tr><td>Available Memory [MB] :</td><td id="Mem_free">-------</td></tr>

2. Register the Node Path in the Map
   Search for text #2# and uncomment the text shown below.
   Currently, two arguments are stored in nodeToRef:
   * ref: reference to the HTML output element
   * digits: number of digits to display
   To implement unit conversion, for example, you can add a suitable parameter here, which then needs to be interpreted in function setValues(). Search for #3#

   <!-- #2# -->
   nodeToRef.set("framework/metrics/system/memavailable-mb", { ref: document.getElementById("Mem_free"), digits: 2 });

3. Open HTML file and follow the displayed instructions

   

Public API Reference of JS classes

Detailed parameters and type structures are declared in JSDoc within the source code. Here is the big picture:

cxApi (Session & REST Manager)

• startSession(...): Establishes session, validates and automatically manages token caching in browser sessionStorage.

• ping(...): Fast, lightweight availability check.

  Dev Tip: Essential to quickly verify if your connection is blocked by CORS/SSL, or if your ctrlX COREvirtual shut down its services (which happens automatically after 4 hours).

• myFetch(...): Robust fetch helper. Intercepts HTTP 401 (Unauthorized) errors, silently re-authenticates via startSession, and automatically retries the original request once.

• stopSession(): Destroys active session on the controller and wipes browser session storage.

cxSubscription (Event Client)

• start(...): Subscribes to the data layer paths and processes the SSE stream.

• stop(): Safely closes the active stream and releases reader locks.

Integrate Subscription in web based visualization

If the subscriptions are used in a web based visualization, different security restrictions apply compared to our test environment. This applies both to the use of iframes and to usage within JavaScript.

The IFrame Security Hurdles

Embedding your local HMI in an <iframe> forces the browser to apply strict boundaries:

• Cross-Origin Mismatch (CORS): If the parent portal (e.g., https://my-hmi-portal.local) and your embedded HMI (e.g., file:///index.html) run on different origins, the browser blocks direct background communication to the ctrlX OS REST API.

• Sandbox Restrictions: Many frame containers use the <iframe sandbox="..."> attribute. If crucial sandbox bypass flags such as allow-scripts or allow-same-origin are missing, Javascript runtime execution and storage access (sessionStorage) are completely disabled.

Direct Controller Hosting

The most robust solution is hosting your HMI files directly on the ctrlX OS internal Web Server.
See article: HowTo view content from multiple web servers in one web page (reverse proxy)

• Why it works: The HMI runs on the exact same origin as the API.

• The Benefit: Direct "Same-Origin" scenario. CORS blocks are bypassed, SSL credentials are automatically trusted because you are already logged in to the control portal, and the IFrame has native access to the browsers sessionStorage with zero security exceptions required

• WebIQ and HELIO use the ctrlX OS reverse proxy.

• CoDeSys WebVisu does not yet use the reverse proxy, but it will soon.

Background information

SSE protocol used by ctrlX OS for subscriptions

• Server-Sent Events (SSE): A unidirectional, standard HTTP/HTTPS-based streaming protocol. ctrlX OS pushes values only on-change, eliminating continuous HTTP-polling overhead.

• Metadata Filtering: The raw network stream includes comment lines (such as :keep-alive sent to prevent browser timeouts) and double newline dividers (\n\n). Our backend client parses and discards them dynamically.

• Here an example of JSON Data Schema sent from ctrlX OS. Each valid update is sent on a data: line carrying a JSON string:

event: update
id: 2
data: {"node":"framework/metrics/system/cpu-utilisation-percent","timestamp":134320390258319784,"type":"double","value":3.9}

event: update
id: 3
data: {"node":"framework/metrics/system/memavailable-mb","timestamp":134320390258321022,"type":"double","value":1298.07421875}

Related Links

• Using REST API of ctrlX CORE

• How to make an HTTPS connection with ctrlX OS Web Server

• How to subscribe to nodes of the ctrlX Data Layer using REST commands via CURL

• Online documentation:

  • ctrlX OS, Operating System for ctrlX CORE Control Devices Version 3, Application Manual --> ctrlX Data Layer --> Subscriptions

  • Subscription - Language binding

Download

I have a project where we need to communicate with another PLC via ethernet IP (Keyence <-> CtrlX X3). I have used the things above (https://community.boschrexroth.com/ctrlx-automation-how-tos-qmglrz33/post/example-for-ethernet-ip-communication-between-ctrlx-core-adapter-and-a-9whD2yQJiHhEK1v?highlight=q4Fc4rxif0kPVq6)

to implement on the CtrlX side. Everything is working fine, but it have happened multiple times that for some reason the communication stopped, and I was only able to restart it by switch off and on the device. Is there any way to debug the ethernet IP connection? The variables that are auto generated for the Ethernet/EThernet IP adater / Ethernet IP module are not very descriptive, and the reset bit does not help at all.

Also, is there any reason to use the ethernet IP app for the CORE (if there is any)?

Thanks in advance!

Today is the day that the new line we have been working on is officially started, and of course, today the system started showing weird behaviors.

Last I have connected to these PLCs was 2026.08.14, back then everything was working fine. Today, I have logged in, but have made no changes. I have also connected to the PLC with WEBIQ designer, to download the localization files, so I can translate stuff.

While this translation was happening, the one of the X3 stopped. I have seen the following in its log:

and on the webserver the following:

On this screen, below the Actions there was a button to Reset Warm/Cold/Origin. Warm was gray, so I pressed reset Cold, then restarted the system. It did not solve the problem, I got the same error message. After that I pressed Reset Origin, and it was working fine.

I also saw the following in the log on the webserver:

How can this happen? What does these lines in the PLC Engineering PLC log means? It is very urgent to find out what caused this problem! Thank you very much!

I also cannot access the Ethercat Master or the Data Layer (Error: 502). If I press "refresh" at the EtherCAT Master page, Error 404 appears.

I checked the dependancies of the EtherCAT Master, too.

With IO Engineering, I can find my PLC but I don't have full access.

It seems, that I have an network error. My PC is not directly connected to the PLC's XF10, it is in another subnet with gateways inbetween. The configuration is managed by a DHCP-server, the IPV4-adress of the PLC is fixed in this server.

The machine had a VPP hardware fault which was fixed with a board level repair. The customer had to re-download the PLC to the CMP60 but now the spindle speed is not being transferred from the "S" command to the PLC variable MD7024_WZ_S_VALUE_Ch1.realVal

I was able to force the value of the PLC variable to another number, eg 10,000. Then I released the force and the value stays the same.

When they run the spindle (eg. run CNC code: M3 S18000) on the CNC the value of the PLC variable goes to zero.

It looks like something is switched off or not configured properly. Can someone make any suggestions what that might be?

I have an archive of the system. I can PM or email it privately for someone to have a look at.

"S" Code Auxilliary function setup below.

• IEC 61850 — reading data from devices

• IEC 60870-5-104 server — exposing that data to a control center

We found two relevant CODESYS Store packages and wanted to confirm platform support before purchasing:

1. IEC 60870-5-104 Server (3S‑Systems, art. 2312000014, €240) — lists "CODESYS Runtime SL Systems" as supported platform (generic). Can anyone confirm this runs correctly on ctrlX CORE? (There's a Bosch community how-to already showing this configured on ctrlX CORE, so this one looks likely fine — just want to double-check current versions.)

2. IEC61850 Client (WAGO, art. 2312000030, €1,299) — its listed "Supported devices" field says "WAGO Edge Computer" specifically, and support is routed to WAGO rather than 3S‑Systems/Bosch. Has anyone successfully installed/run this on a Bosch ctrlX CORE? Or is it locked to WAGO hardware (via licensing, feature dependency, or vendor certification)?

Specific questions:

• Is IEC61850 Client (2312000030) known to work on ctrlX CORE's CODESYS Control SL runtime, or is it restricted to WAGO controllers only?

• Is there a Bosch-recommended/alternative IEC 61850 MMS client package for ctrlX CORE that we should use instead?

• For either package: is the license a one-time perpetual license per device (soft container), or a recurring/subscription cost? We're planning for multi-site rollout, so this affects cost projections significantly.

• Any known performance/stability caveats running both add-ons simultaneously in one Automation Runtime application?

nickH CodeShepherd

We have been working for some time on a CtrlX Core application with a fairly diverse EtherCAT architecture, including XMD drives, a SAFEX C12 safety PLC, IO-Link master, I/O node, a pneumatic valve island, etc.

We are now at a fairly advanced stage of the project, and the configuration has been validated for some time.

However, we have recently encountered an unexpected issue during startup. When we power on the machine, everything would normally start up without any problems. But now, the CPU does not start correctly and remains in a partially operational state, while the EtherCAT network remains in Pre-OP.

Here is the EtherCAT diagnostic information I am getting:

I do not understand why the EtherCAT bus does not start.

Could you please provide me with some guidance on how to investigate this issue and identify the root cause? In order to start the system, I manually switch the EtherCAT network via the web server from Init to OP. At that point, everything starts up normally.

Hardware and Software:

Thank you in advance for your help.
Hasan