资料简介：
These events need to be analyzed and presented as reports to decision makers. Consider a
typical use case of generating aggregate information. For example:
• What were the sales of a product at various intervals (every 5 minutes, 15 minutes, 30
minutes, 1 hour, 24 hours, etc.)?
• Did the sales for a product increase in the day following the introduction of a promotion
scheme?
• How many hits did the page generate in the first 1 hour of being published?
• What are the key words associated with the user activity on our news website in the
last 5 seconds? Which ads do we show to this user based on this information?
The most common method of implementing such solutions is to collect the daily transactions
in a staging database and perform aggregations and other analytics operations in a batch
processing mode. The output of this batch processing system is then collected in a Data
