Hello! В сообщении от Wednesday 19 November 2008 15:05:01 D. Richard Hipp написал(а): > That's because "SELECT count(*) FROM t" has to read and count every > row in the table - all 1.2 million rows. And this involves reading > all 1.2 million rows from disk. If each row is 100 bytes in size, > that involves reading 120MB.
Is it possible using only PRIMARY KEY index for select count(*) queries? For other indexes it's possible as explain query plan select count(*) from street where code >= '77000000000' and code < '77000000000'+1; 0|0|TABLE street WITH INDEX street_code_idx Best regards, Alexey. _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
