I do not see anything obviously wrong with your code, although of course I could not analyze it in full, just "scan it" on its surface. One thing that somewhat puzzles me is the use of Mutex which looks heavyweight here, if it serves just to work as a C# "lock" statement (a critical section), plus the way it is coded is not safe (exceptions will leave the mutex in an incorrect state) - but unless you see the code being deadlocked somewhere, this should not be related to the problem reported.

Therefore I do not currently know what the reason for the errors can be. One possible thing to do involves some extra work, and it would be to deploy an OPC Analyzer (which we can provide to you) to monitor the communication between the client and the server. Doing so should at least allow deciding with reasonable certainty whether the problem is on the client side or the server side. However the OPC analyzer acts as a middle OPC client/server piece in the communication therefore it needs a modification in your client app (to point to a different server), and can sometime introduce new issues. Let me know how you would like to proceed.

Best regards

To me, that looks more like an accident rather than a solution.

Can you please answer the following questions:

1. Is your app multi-threaded?
2. Are you creating more than one EasyDAClient instance?
3. Can you describe in general items what the app is doing with regard to EasyDAClient calls (the "usage pattern") - the sequence of methods it executes.
4. When you wrote "I specify that reading and writing multiple items are only made on running PLC without bad quality items, and not on the stoped PLC." - how is that achieved? Do you sequentially first call ReadMultipleItems for items that reside on one PLC, and then for a second PLC, or is there a different approach?

Thank you

Have you considered that this may be a genuine timeout, i.e. that the server truly takes more than 5 seconds to fulfill the request? Remember that it has to work on multiple items at once, and can only return the results when it has the data for all of them, and (depending on the implementation of the server) in the case of PLC down, it may take longer than usual - a time is needed to determine that there is a communication problem.

What happens if you increase the timeout significantly?

Best regards