Hi??
I've already read the development manul about these three methods. And it says that putall is inefficient when cluster loads a large nember of data. Here comes my problems. What are the advantages of localcache and streamer in big data? Are there any mesures to ensure data is not lost? Like at least once or only once...
