mysql - Fast selection of a block of following records -
is there implementation of sql database allowing select table block of n following records -- in respect of index -- starting specified index, performance o(n + log tables_size)
? allowing adding record in o(log tables_size)
. if so, how it?
i'm dreamer but, possible mysql?
if id
primary key on table, following return in order of time needed fetch records, plus initial index seek:
select t.* t id >= someid order id limit <n>;
adding record consists of 2 parts. first part finding available space , second part inserting index. b-tree index should require o(log table_size)
insert. if pages full , inserting @ end of table, finding right page constant time.
in other words, if understand question correctly, primary key clustered indexes asking for.
Comments
Post a Comment