OPC Client with Yokogawa Centum Server

[mffba]

Dear expertise, 

First of all, thank you for spending your time to read this.Let me brief a little bit about my project.

3 types of PCs were setup in a training labarotary which are LabVIEW PC, lecturer PC and Student PC. At LabVIEW PC, We’d developed a LabVIEW program to work as an OPC Client to 3rd Party OPC Server, Yokogawa Centum. Since the communication to this 3rd party server is not very straight forward, DataSocket was implemented as it allows easy data transfer over many different protocols.

At lecturer PC, we’d deployed network-published shared variables to a shared variable engine (SVE) that hosts the shared variable values on the network. The SVE works as server for a shared variable and all the references (IO tags from Yokogawa Centum) are the clients.

SVE at Lecturer PC published the 3rd party IO tags meanwhile DataSocket at LabVIEW PC subscribes it. The OPC Server will response to the feedback from LabVIEW program via DataSocket and distributes the signal to 24 Student PCs IO tags accordingly.

The issues only came in when we tried to deploy more than 500 tags at lecturer PC

1000 I/O deployment

LabVIEW software unable to deploy more than 500 I/O. When more than 500 I/O deployment has done, issues occur such like all the tag will crashed and disconnected from the 3rd Party OPC Server (refer to Distributed System Manager screenshot). Even though we find out we need to wait for around 45 minutes in order to complete the connection (during this period all the I/O is  not connected) but we still  encounter other issues like  LabVIEW unable to WRITE data into our OPC.

Delaying of communication

There some communication delay between the LabVIEW deployed tag and 3rd party OPC server is averagely more than 2 seconds and maximum can reach 8 seconds. Need to advised on how to reduce the delaying of the deployed tag to our OPC server. Delay is happening when we deploy up to 500 I/O, I believe the delaying time is proportional to the amount of I/O  that we deploy.

I was thinking to upgrade the lecturer PC spec to slightly higher since the 3rd party server and our Publisher (SVE) running at the same PC and they might  have race condition on the resource. However,I cannot guarantee whether these really could solve my issues. Appreciate if someone who has the knowledge or similiar experience mind to share their tought on how can I solve this and do advice me what might be the best solution. Thanks in advance

[bruno_costa]

I think your problem is with the DCOM security.
The OPC server has two operating modes: Synchronous and Asynchronous.
The Synchronous mode is slower than the asynchronous mode.
When there are problems in the security configuration of DCOM OPC client can not establish an asynchronous connection  and tries to establish a synchronous connection, which is slower.

To help solve this problem I recommend you a software called OPC Rescue, which can be found in www.opcti.com for free.