Indexing criteria: disk space usage

The following table provides estimates of the amount of space each index uses compared to the amount of column data from the source database or flat file.

Table 6-3: Index disk space usage

Type of index

Estimated space versus raw data

Comments

Default

Smaller than or equal to

If the number of distinct values is less than 255, this index uses significantly less space than the raw data

High_Group

Smaller than up to 2 times larger

As the number of distinct values decreases (that is, the number of entries per group increases), the space used decreases in proportion to the size of the raw data

High_Non_Group

Smaller than or equal to

Smaller than the raw data in most cases

Low_Fast

Smaller than up to 2 times larger

Same as High_Group

Date

Smaller than or equal to

Larger than High_Non_Group

Time

Smaller than or equal to

Larger than High_Non_Group

Datetime

Smaller than or equal to

Larger than High_Non_Group

For LF and HG indexes, the index size depends on the number of unique values. The more unique values, the more space the index takes.

Because CMP indexes are always an additional index, they do not save disk space.