Fragmentation and aggregation control method for storing big data under high concurrency

The invention discloses a fragmentation and aggregation control method for storing big data under high concurrency, which comprises the following steps of: configuring a platform, and configuring a fragmentation unit of an index needing to be fragmented, wherein fragmentation is rounding calculation, each piece of data has a timestamp with a service attribute when entering statistics, and the timestamp is data of a Long type; then defining segment units as milliseconds, seconds, minutes, hours and days, so that the duration corresponding to the time units are respectively 1, 1000, 60000, 360000 and 86400000, and then, the duration = parameter entering timestamp/segment units, so that each duration is a part of the whole time span, and the duration does not intersect with each other. According to the method, the problem of large value storage of a non-relational database is solved; preliminary aggregation calculation is carried out, later access calculation is reduced, and meanwhile the data storage space is r