Question

I am trying to make multithread use of one sqlite database on Android.
I have the following situation :

Activity A : it has 10 threads which use DB ,but uses only one SQLiteDatabase

Activity B : has 10 threads which use DB ,but uses only one SQLiteDatabase

(the two SQLiteDatabase objects of the Activities are different)

The logic execution is the following :

1. A calls B
2. B inserts into the database and returns to A
3. A display the result
4. A try to erase what B inserted but fails .

A can erase old things ,but the new objects that B just inserted fail,if I restart the application ,A is able to erase the objects that previously couldn’t.
Should I use the same SQLiteDatabase for the two Activities ,and if is this the case how to pass this object to Activity B ?

Activity A thread code :

private class Delete_Task extends AsyncTask {

    private Context mContext;

    private DatabaseManager mDB;

    public DeleteEvent_Task(Context mContext,DatabaseManager DB) {
        this.mContext = mContext;
        this.mDB = DB;
    }

    @Override
    protected Void doInBackground(Long... params) {
        long id = params[0];

        this.mDB.delete(id);

        return null;
    }
}

Activity B thread code :

private class Update_Task extends AsyncTask< Object, Void, Void > {

    private Context mContext;


    public Update_Task(Context mContext)
    {
        this.mContext = mContext;
    }

    @Override
    protected Void doInBackground(Event... params) {
        Object e = params[0];

        DatabaseManager DB = new DatabaseManager(this.mContext);

        if (e.getId() <= 0)
        {           
            DB.add(e);
        }
        else
        {

            DB.update(e.getId(), e);
        }
        DB.close();

        return null;
    }       
}

Answer 1

By far the simplest thing to do is open the database connection once, on application startup, and share this connection throughout your code without ever closing it. Provided you use a single connection, all reads and writes will be correctly serialized so you don’t need to worry about it.