Multithreading – Multithreaded File Copy in .NET

This is going to be a large and very involved project. I've been responsible for migrating Access databases to .NET many times, but never on this scale. I agree that a gradual approach will probably be the most realistic. Trying to take on everything in one shot is an easy way to fail.

As in other answers, the first and most important step will be to move everything to SQL Server. While doing this, document the database if this hasn't already been done, and identify areas for refactoring if they exist. Access databases that grow to meet evolving needs often have data models that can be improved upon when looking at the big picture.

Next, identify modules of the application that are "self contained". Because this is a do-everything database, there will probably be modules of it that are nearly completely disjoint from each other. After having these disjoint pieces, identify one of the smaller modules that has the most to gain from being updated, and take it on first.

While migrating to .NET, don't just do a simple 1 to 1 copy of the application. Gather input from users to identify pain points with the current application. You want your first migration unit to be a huge success to gain the complete buy in from everyone else.

After you have the first unit done, look at what happened in terms of challenges, difficulties, etc, and use this moving forward.

One thing that I would strongly discourage would be to implement partially working modules (ex: "we migrated the CRM, but features X,Y,Z are not in the new system yet"). This will cause users to quickly become frustrated from having an incomplete system when the old one to them was "perfectly fine". This kind of development may work fine for other domains, but not in businesses where users see "nothing wrong" with the old Access monolith and don't understand the migration needs.

Your code has significant other issues apart from just that. Manually deleting a pointer? Calling a cleanup function? Owch. Also, as accurately pointed out in the question comment, you don't use RAII for your lock, which is another fairly epic fail and guarantees that when DoSomethingImportant throws an exception, terrible things happen.

The fact that this multithreaded bug is occurring is just a symptom of the core problem- your code has extremely bad semantics in any threading situation and you're using completely unreliable tools and ex-idioms. If I were you, I'd be amazed that it functions with a single thread, let alone more.

Common Mistake #3 - Even though the objects are reference counted, the shutdown sequence "releases" it's pointer. But forgets to wait for the thread that is still running to release it's instance. As such, components are shutdown cleanly, then spurious or late callbacks are invoked on an object in an state not expecting any more calls.

The whole point of reference counting is that the thread has already released it's instance. Because if not, then it cannot be destroyed because the thread still has a reference.

Use std::shared_ptr. When all threads have released (and nobody, therefore, can be calling the function, as they have no pointer to it), then the destructor is called. This is guaranteed safe.

Secondly, use a real threading library, like Intel's Thread Building Blocks or Microsoft's Parallel Patterns Library. Writing your own is time-consuming and unreliable and your code is full of threading details which it doesn't need. Doing your own locks is just as bad as doing your own memory management. They have already implemented many general-purpose very useful threading idioms which work correctly for your use.