Who wrote the server code that is trying to access the device?

This sounds like a bug in the code that provides access to the device.

I am not sure whether this is an issue with our server code or not. We have written our OPC UA server based on OPC UA Sample server.

All I wanted to know this.

Two OPC UA clients connected to one server, clients wants to read data simultaneously. Client 1 would read 30 items which would take 1 minute and 30 seconds to complete. Client 2 would read 40 items which would take 1 minute and 40 seconds to complete. When both the client issues read calls to server, is server handle them using two different thread(like schedulers) or is server using a single thread to complete Client 1 read calls and then process client 2 calls. Latency for Client 2 is too high if we increase the number of items/ or if we use lower baud rate.

I want OPC UA server or our own scheduler to handle them using different threads instead of using a single thread and queue model for client requests. If the OPC UA framework, didn't do this scheduling, there might be a place where we can write our own schedulers.

I might be ENTIRELY wrong in this approach. If then, please show me the right path.