Advanced MySQL 8

A typical use case: time series data

As already mentioned, partitions are commonly used when dealing with a dataset containing many records. These may include phone call records, invoices, subscriptions, and so on.

In general, we all want our reports to present new data, and unfortunately there will be old data to delete once they are out of date and not at all useful.

To illustrate how to deal with this problem, I will present the scenario of a company that uses two million records for a gas station network. We will be performing some common operations on the metrics table as it-is, and partitioning a copy of the table by RANGE to identify any potential differences.

The following example will have millions of records and two tables: