Stonebraker about the database five key

Michael Stonebraker is a pioneer in the field of databases. His deep expertise has also helped to develop some of the most popular database systems such as Postgres, Ingres and Vertica. One of his most recent studies is VoltDB, a memory-based OLTP (Online Transaction Processing) system, which he claims is two orders of magnitude faster than traditional solutions.

However, Stonebraker's views have been controversial over the years, with strong support on the one hand and strong opposition on the other. For example, he said in 2011, Facebook is stuck in MySQL, "life is better than death"; one can imagine, come back quickly and fiercely.

Stonebraker visited this week's Structure Show, talking about his assessment of the database market, including NoSQL, the fate of Oracle, of course, Facebook's MySQL issues, here are some highlights:

A pattern will not fit all

"In any conceivable vertical market, there is always a better solution than a traditional relational database system," Stonebraker said.

In fact, he has always promoted this idea, but now it seems more reasonable than before. There are column storage schemas useful for analysis, memory schemas for transactions, and of course NoSQL architecture for key-value operations and new datatypes, and even graph databases are in the pipeline.

To prove how these new types of databases are becoming mainstream, Stonebraker said: "Like Obamacare, good or bad, has been built on NoSQL"

There can be multiple winners in this area

"There will be three or six different types of database system architectures with two or three successful vendors at every level," Stonebraker predicts. "I think the most important thing is that the traditional relational database system will slowly shrink, and all this may happen in a decade."

NoSQL's strong ground

Stonebraker said: "NoSQL will mean more than SQL, and Cassandra and MongoDB have confirmed my prediction." For example, the purely low-level language has little value, Stonebraker that NoSQL system will embrace ACID in the future, and all this is already occur. Stonebraker thinks

NoSQL non-ACID biggest supporter is undoubtedly Google's Jeff Dean, he is basically responsible for all Google's database products. Recently, he wrote a system called Spanner, arguably a pure ACID system. Therefore, Google is also moving towards ACID, meanwhile, NoSQL market will also get rid of the final consistency toward ACID.

Oracle will feel the pressure from SAP

Among other things, I think it's really interesting that SAP has not got the attention it deserves in the database space, and now SAP's customers are also Oracle's biggest customers. Among these giants, Oracle and SAP will compete face to face.

Stonebraker argues it may be a bit early to happen, and it's not yet known how SAP's customers will respond under constant lobbying, but SAP is best positioned to give its customers a reason to move from Oracle to HANA .

Facebook will continue to look for a MySQL alternative, but it may not work

Facebook has become the toughest data management issue on the planet, and MySQL has been trying for years, but so far no database systems have been found that match their size.

Stonebraker pointed out that even if MySQL in the Facebook efforts to match the size of the social giant business, but this is obviously not a positive affirmation of MySQL, but Facebook's accomplishments in the field of database praise.

In general, Stonebraker said (back to familiar but interesting wording) that "traditional code base vendors are 25 years old and it's time for them to retire."