Well, there are three types of databases: Relational, Columnar, and finally massively distributed. They throw away different feature sets in order to handle a certain subset better.
The primary reason Relational databases are so friggin' strong is that they have a solid foundation and maturity. Most new "NoSQL" stuff is immature crap which should never be used in production for any kind of persistent data. But they are being used as such, and I do note that matures the products over time.
What is the rationale of putting this work into Mysql? With its sloppy execution and the dark clouds of Oracle hanging over it? Why not Postgres, or even MariaDB instead?
Q: Why didn't you base this on MariaDB, Percona Server, Drizzle, etc....
A: We reached a consensus that MySQL-5.6 was the right choice for this, as it has the production-ready features we need to operate at scale, and the features planned for MySQL-5.7 seem like a fitting path forward for us. We will continue to revisit this decision as the ecosystem evolves.
[+] [-] collyw|12 years ago|reply
[+] [-] jlouis|12 years ago|reply
The primary reason Relational databases are so friggin' strong is that they have a solid foundation and maturity. Most new "NoSQL" stuff is immature crap which should never be used in production for any kind of persistent data. But they are being used as such, and I do note that matures the products over time.
[+] [-] ams6110|12 years ago|reply
[+] [-] avz|12 years ago|reply
- they provide unnecessary features at high cost (e.g. database-wide transactions),
- they don't provide features that are essential (e.g. scalability, distribution, graceful degradation).
By contrast NoSQL data stores like Bigtable are highly scalable and can easily be stacked over to provide more features if needed (see e.g. Megastore, http://pdos.csail.mit.edu/6.824-2011/papers/jbaker-megastore...)
[+] [-] mixmastamyk|12 years ago|reply
[+] [-] morgo|12 years ago|reply
Also from the WebScaleSQL FAQ:
Q: Why didn't you base this on MariaDB, Percona Server, Drizzle, etc.... A: We reached a consensus that MySQL-5.6 was the right choice for this, as it has the production-ready features we need to operate at scale, and the features planned for MySQL-5.7 seem like a fitting path forward for us. We will continue to revisit this decision as the ecosystem evolves.
[+] [-] nl|12 years ago|reply
[+] [-] davidw|12 years ago|reply
[+] [-] rakoo|12 years ago|reply
[+] [-] peferron|12 years ago|reply
Q: Why is it called WebScaleSQL? A: While there are a variety of origin stories for the name depending on who you ask, ...
They just don't want to mention it directly. :)
[+] [-] RyanMcGreal|12 years ago|reply
http://www.youtube.com/watch?v=b2F-DItXtZs
[+] [-] jpalomaki|12 years ago|reply
Project page: http://webscalesql.org/
[+] [-] mikkelewis|12 years ago|reply
[+] [-] arrc|12 years ago|reply