this is a common performance measure in data science. It is not an error in the way of "there went something wrong" but rather a goodness of fit. The lower this value, the better is your model. Just google for it with "regression" and you find plenty of ressources.
So is that means there is no problems with my opration, but my data fit not that well in the model?
There is indeed no problem. Your model might be perfect. Every model has an RMSE. You can use this to compare it to another model. The smaller the better.
RapidMiner Auto Model
RapidMiner Turbo Prep
Reports & Tools
Educational License Program