We are trying to use the 'in process' version of java db4o in our application. Our app produces lots of data pretty frequently and we want to use a background thread which keeps taking chunks of data and saving it to a db4o database.
My question is that if we use multiple threads to store chunks of data parallely will that increase the performance at all?
My ambiguity is primarily that since the disk is only 1 and thus can be accessed by a single thread at a time, will using multiple threads make any difference in the performance?
For example, lets say our app produces 100 mb of data. Would it make any difference if we split it into 2 chunks of 50 mb and use 2 different threads to save the data?