OPC COM Call Tracing

OPC COM Call Trace Technology

Diagnose OPC Problems Fast

The root cause of OPC interoperability problems is sometimes difficult to determine. When you have a system to commission and put online, you want answers fast. That is why all Northern Dynamic products include our OPC COM Call Tracing Technology.

OPC COM Call Tracing is like having a built-in protocol analyzer for your OPC call stack.

When activated, OPC COM Call Tracing creates an unbiased, detailed record of each OPC COM call made between an OPC client and an OPC server. Using this record, our OPC Experts will diagnose and report the root cause of the problem.

How To Try OPC COM Call Tracing

Download and install one our OPC server toolkit product evaluations. Each evaluation installs one or more sample OPC servers. Run the server and activate the OPC COM Call Tracing feature. Connect to the server with your client application and interact with the server, browsing or creating groups and items.

Now look under the directory C:\Program Files\Common Files\Northern Dynamic\NDITraceHook for the file NDITraceHook.log. Complete OPC call stack information is logged. Read on for more details.

Click here to view excerpts from a sample log file.

How OPC COM Call Tracing Works

OPC COM Call Tracing is a software component that records OPC COM method calls between the application that contains the component and other OPC compliant applications.

For example, the OPC COM Call Tracing component in the OPC Gateway (an OPC client) records call details between the gateway and OPC servers to which it connects.

Similarly, an OPC server created with one of our OPC Server Toolkits can record call details between the server and OPC clients that connect to the server.

TraceHook Software Component

OPC COM Call Tracing is implemented by the NDI TraceHook Component, a DLL that hooks into OPC COM calls to capture and record the call stack. Call stack information is decoded, annotated, and logged to a circular file. The diagram illustrates how the TraceHook component is integrated with the OPC DA Server Toolkit.

Since the TraceHook component hooks the COM calls (and does not operate in a pass-through mode) OPC COM Call Tracing is very efficient, imposing very little run-time overhead. Therefore, timing related integration problems can be diagnosed.

The TraceHook component ships with a simple utility application that allows tracing parameters to be changed dynamically.

Logging Details

OPC COM Call Tracing logs timestamped, thread-specific details for all OPC DA interface method calls. Developers use this tool to create an exhaustive and annotated audit trail of OPC interface calls on your server. When activated, the following information is logged for each OPC interface call.

Pre- or Post-method call timestamp.

Unique COM object ID for identifying the object instance making the call.

Unique, sequential call ID for identifying the order of calls on a per thread basis.

Interface name and method name.

HRESULT for post-call messages.

[in] and [out] parameter values.

The following levels of COM call details can be selected for logging.

Only failed method calls.

All method calls with HRESULTs only.

Pre- and post- method call trace messages to calculate call timing.

Pre- and post- method call trace messages with full [in] and [out] parameter dump.

For all levels, a filter expression can be specified. You can dynamically filter the message stream, reducing the amount of information logged to that which is of interest.