Database storage overhead

diff -r 000000000000 -r 89d6a7a84779 Symbian3/SDK/Source/GUID-785B2F0B-E7E6-5DAE-98F1-6C32BED25964.dita --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/Symbian3/SDK/Source/GUID-785B2F0B-E7E6-5DAE-98F1-6C32BED25964.dita Thu Jan 21 18:18:20 2010 +0000 @@ -0,0 +1,12 @@ + + + + + +Database storage overhead

Space utilization is efficient

the underlying permanent file store adds minimal overhead to maintain its structure: store imposes a fixed 46 bytes + 2 bytes per additional 16K; each stream requires 7 bytes.
the database has minimal space requirements to store its schema and data structure.
including the stream overhead, row storage overhead can be less than 2 bytes per row in ideal clustering conditions, even with lower clustering the overhead is usually below 1% of the data volume.

The storage required for a row can be determined as follows:

each non-null value requires the storage for fixed width columns, and the raw storage plus 1 byte for variable width columns.
long columns are stored embedded in the row data when they are small enough, and otherwise they are stored in a separate stream: this makes more efficient use of the both space and speed. When embedded they require just 1 bit more than the short columns, when separated they require 8 bytes in the row data plus any stream overhead.
each nullable column requires an extra 1 bit.
bits are packed into bytes in the row storage.

Indexes

Indexes are implemented using STORE B+trees. Note in particular that indexes use fixed length keys, so that indexes on longer text fields can consume significant space.

If the key for the index is k bytes, the number of rows to index is n, the index page size is P, and the B-tree packing density is r:

a = [(P-8)/(k+4)] * r

Where [x] is the largest integer <= x. Then the number of pages required, N, is

N = n * (1/a + 1/(a*a))

Each page requires P+7 bytes in the store, so the total indexing overhead, S, is

S = N * (P+7)

For DBMS P=512 and r=0.86.

\ No newline at end of file