Google goes back to the future with SQL F1 database by Jack Clark.
From the post:
The tech world is turning back toward SQL, bringing to a close a possibly misspent half-decade in which startups courted developers with promises of infinite scalability and the finest imitation-Google tools available, and companies found themselves exposed to unstable data and poor guarantees.
The shift has been going on quietly for some time, and tech leader Google has been tussling with the drawbacks of non-relational and non ACID-compliant systems for years. That struggle has demanded the creation of a new system to handle data at scale, and on Tuesday at the Very Large Data Base (VLDB) conference, Google delivered a paper outlining its much-discussed “F1” system, which has replaced MySQL as the distributed heart of the company’s hugely lucrative AdWords platform.
The AdWords system includes “100s of applications and 1000s of users,” which all share a database over 100TB serving up “hundreds of thousands of requests per second, and runs SQL queries that scan tens of trillions of data rows per day,” Google said. And it’s got five nines of availability.
(…)
F1 uses some of Google’s most advanced technologies, such as BigTable and the planet-spanning “Spanner” database, which F1 servers are co-located with for optimum use. Google describes it as a “a hybrid, combining the best aspects of traditional relational databases and scalable NoSQL systems”.
(…)
I am wondering what the “…RDBMS doesn’t do X well parrots…” are going to say now?
The authors admit up front “trade-offs and sacrifices” were made. But when you meet your requirements while processing trillions of rows of data daily, you are entitled to “trade-offs and sacrifices.”
A very deep paper that will require background reading for most of us.
Looking forward to it.