As much as the number of data increases, it’s important to think “HOW” to manage thousands of files in cloud storage. In service provider’s point of view, their concern is to think how to effectively manage a big amount of data and improve services. On the other hand, users think of how to save my files and use it conveniently.

As for today, let’s focus on the service provider’s point of view and how they try hard to manage system smoothly.

Hot data? Cold data?

What should cloud storage provider need to do in order to provide sufficient storage space and fast speed? It’s obvious that they need to build more storage spaces and use fast storage and create efficient network services. Thus, people most often create a data center called IDC to provide cloud service to users. However, it costs a lot to retain and difficult to build storage spaces time after time.

In order to reduce costs and create effective atmosphere, service providers analyze the data which are stored in cloud storage. According to the paper, 70~80% of data in network rarely accessed or used by the user. Data that is accessed frequently are about 20~30% and those data become accessed less frequently as time goes by. This often referred to as the temperature, hot, warm, and cold. Hot data is data which is accessed very frequently and warm data is accessed less frequently than hot data. Cold data is data which is either never accessed or accessed very rarely.

Most of the cloud storage saved data are cold data, so managing those data can increase system efficiency of cloud storage or system. Therefore, most cloud service providers divide data into cold or hot data and use fast speed storage for hot data, and use slower storage for cold data. Facebook uses a customized data center to manage 2 billion of pictures of cold data.

Stay tuned for more tips and information on our next blog posting! 🙂



Please enter your comment!
Please enter your name here