2X-20X is hardly "small...overhead" in my world. Even 2X is the difference between 30 days and 15 days. One 16-computer blade rack vs two racks ($200,000 vs $400,000).
That's why google did this. Works for what they need and is lots cheaper. Michael D. Black Senior Scientist NG Information Systems Advanced Analytics Directorate ________________________________ From: sqlite-users-boun...@sqlite.org [sqlite-users-boun...@sqlite.org] on behalf of Alexey Pechnikov [pechni...@mobigroup.ru] Sent: Friday, July 29, 2011 7:44 AM To: General Discussion of SQLite Database Subject: EXT :Re: [sqlite] LevelDB benchmark 2011/7/29 Black, Michael (IS) <michael.bla...@ngc.com>: > What they don't say explicitly is that if all you need is key/value > capability then an SQL database is overkill and only slows you down (bit of a > duh factor there though not obvious to neophytes). The overhead by SQL layer is small. And are used prepared statements in test. As I see LevelDB use data integrity equal to SQLite WAL mode with disabled fsync. And LevelDB is limited only single-thread access. With the patched test SQLite is about 2x - 20x slower and it's absolutly normal I think. With dirrefent page_size we can make some tests faster. P.S. There is constant database created by DJB. And exists patch to drop "constant" limitation. IMHO it's functionally equal and better solution than LevelDB... -- Best regards, Alexey Pechnikov. http://pechnikov.tel/ _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users