Compacts the internal representation of the concurrent vector to reduce fragmentation and optimize memory usage. This method is not concurrency-safe.
This method will internally re-allocate memory move elements around, invalidating all the iterators. shrink_to_fit is not concurrency-safe. You must ensure that no other threads are invoking methods on the concurrent vector when you call this function.