Computing column access statistics in MySQL and / or Oracle -
We currently have your company's data model (database programs), which wildly and distributedly have increased over the past few years And now they are suffering from redundancy and lack of clarity. Developed by different teams for various databanks (Oracle and MySQL), are asked by different tools for different purposes (billing, reporting, business, etc.).
different tools to get a clearer picture of the data dependency, ie which requires tool t pieces of accurate information, in a way that will identify them directly from SQL queries that implement each database . Alternatively, this databank will be more interesting to gather in these statistics.
Such data will provide the
- How often table column TC can be accessed
- There are many T1 time table T2 joined together
Anyone know how to do this with MySQL or Oracle or both?
Oracle has a SYS.col_usage $ table statistics on how different where different columns referenced in section . Unfortunately, there is no record of relationship between columns. Apart from this, no record of any column has been read (like in a WHERE section).
EQUALITY_PREDS column which are a reference in the segment: COL = 123 is EQUIJOIN_PREDS are to meet as A.COL = B.COL there NONEQUIJOIN_PREDS, RANGE_PREDS, LIKE_PREDS, NULL_PREDS Counters for
The data are collected and made available to the system, while the data is being generated.
An example script is here:
Comments
Post a Comment