Efficient processing of data extents

A computer-implemented method for optimization of query processing of data extents, the method comprising:

• storing entries of a table on a set of data extents;
• maintaining attribute value information for each data extent of the set of data extents, the attribute value information comprising a minimum value and a maximum value of an attribute of the entries stored in the respective extent;
• determining a first metric of a first data extent of the set, wherein the first metric utilizes the attribute value information of the first data extent, to determine a split of the first data extent into a first sub-extent, and a second sub-extent, will increase a query processing efficiency;
• splitting the first data extent into a first sub-extent and a second sub-extent based on the determined first metric;
• performing a clustering analysis using the attribute value information of the set of data extents to define the first sub-extent and the second sub-extent;
• confirming the splitting was optimal based on a function calculated for the first data extent, the first sub-extent and the second sub-extent:
• f
• ⁡
• (
• z
• ⁢
• ⁢
• 1
• min
• ,
• z
• ⁢
• ⁢
• 1
• max
• ,
• )
• =
• z
• ⁢
• ⁢
• 1
• max
• -
• z
• ⁢
• ⁢
• 1
• min
• 2
• 32
• ,
• where z1max (z1min) indicates maximum (minimum) values of the attribute value for the first data extent, the first sub-extent, and the second sub-extent, where z1 is a 32 bit integer, wherein the splitting is optimal if a value of the function for the first sub-extent and a value of the function for the second sub-extent are both less than a value of the function for the first data extent;
• determining an attribute value information of the first sub-extent and determining an attribute value information of the second sub-extent;
• determining a second metric of a second data extent of the set and a third data extent of the set, wherein the second metric utilizes the attribute value information of the second data extent and the third data extent, to determine a merge of the second data extent and the third data extent will increase the query processing efficiency;
• merging the second data extent into the third data extent based on the determined second metric;
• and
• updating the attribute value information of the second data extent.