MIT Prof. Michael Stonebraker: “The Traditional RDBMS Wisdom is All Wrong”

A very interesting talk about the future of DBMS was recently given at EPFL by MIT Professor and VoltDB Co-founder and CTO Michael Stonebraker, who also gave us Ingres and Postgres. In a bit less than one hour, he explains his views with respect to the three main pillars of database management systems:

  • OLAP / Data warehouses
  • OLTP
  • Other types of data stores

As a NewSQL vendor also actively involved with H-Store, he is of course heavily yet refreshingly biased towards traditional RDBMS storage models being obsolete (an interesting fact is that Oracle Labs representative Eric Sedlar also attended the talk. One might think that the talk was a slighly FUD-dy move against a VoltDB competitor). Unlike what has come to be known as the NoSQL movement, NewSQL relies on similar relational theory / set theory as “traditional SQL”, including support for ACID and structured data.

His claims mainly include that:

  • OLAP / data warehouses will migrate to column-based data stores within 10 years. The traditional row-based data storage approach is dead, as row-based storage will never match column-based storage’s performance increase by factor 100x.
  • For OLTP, the race for the best data storage designs has not yet been decided, but there is a clear indication of classic models being “plain wrong” (according to Stonebraker), as only 4% of wall-clock time is spent on useful data processing, while the rest is occupied with buffer pools, locking, latching, recovery.
Image from Stonebraker's presentation depicting the amount of "useful" work performed by any RDBMS
Image from Stonebraker’s presentation depicting the amount of “useful” work performed by any RDBMS

I specifically recommend the OLTP part of his talk, as it shows how various new techniques could heavily increase performance of traditional RDBMS already today:

  • Most OLTP systems can afford to buy the amount of memory needed to keep data off the disk. This will remove the need for a buffer pool.
  • Single-threading would get rid of the latching overhead. H-Store and VoltDB statically divide shared memory among the cores, for instance. This is very important as latching gets worse and worse with the increasing amount of cores we have, today.
  • Dynamic locking is not really implemented in any popular RDBMS, but the market is uncertain, which workaround best implements concurrency control. In his opinion, MVCC is not going to do the trick in the long run.
  • ACIDness is something that even Jeff Dean from Google admits to miss, once it’s gone, as eventual consistency does not really keep its promise.
  • In a cluster, active-active consistency management can increase log throughput by factor 3x, compared to active-passive logging. (active-active = transaction is run on every node, active-passive = transaction is run only on the master node, the log is sent to all slave nodes)
  • And also, very importantly, anti-caching is a good technique when the in-memory format matches the disk format, as traditional RDBMS spend a substantial amount of time converting disk data formats (blocks, sectors) into memory formats (actual data).

The essence of Stonebraker’s talk is that the “elephants” who currently dominate the market are too slow to react to all the NewSQL vendors’ innovations. It is a very exciting time for a database professional (some refer to them as data geeks) to enter the market and publish new findings.

Another interesting thing to note is that SQL (call it NewSQL, OldSQL) will remain a dominant language for querying DBMS, both for column-stores as for row-stores. This is a strong statement for tools like jOOQ, which embrace SQL as a first-class citizen among programming languages.

See the complete talk by Michael Stonebraker here:

See Stonebraker's Talk here: http://slideshot.epfl.ch/play/suri_stonebraker
See Stonebraker’s Talk here: http://slideshot.epfl.ch/play/suri_stonebraker

Further reading:

“NoSQL” should be called “SQL with alternative storage models”

Time and again, you’ll find blog posts like this one here telling you the same “truths” about SQL vs. NoSQL:

http://onewebsql.com/blog/no-sql-do-i-really-need-it
(OneWebSQL being a competitor of jOOQ, see a previous article for a comparison)

Usually, those blogs aim for the same arguments being:

  • Performance (“SQL” can “never” scale as much as “NoSQL”)
  • ACID (you don’t always need it)
  • Schemalessness (just store any data)

For some funny reason, all of these ideas have led to the misleading term “NoSQL”, which is interpreted by some as being “no SQL”, by others as being “not only SQL”. But SQL really just means “Structured Query Language”, and it is extremely powerful in terms of expressing relational context. It is well-designed for ad-hoc creation of tuples, records, tables, sets and for mapping them to other projections, reducing them to custom aggregations, etc. Note the terms “map/reduce”, which are often employed by NoSQL evangelists.

For good reasons, the Facebook Query Language (FQL), one of the leading NoSQL query languages, closely resembles SQL although it operates on a completely different data model. Oracle too, has jumped on the “NoSQL” train and sells its own product. It won’t be very long until the two types of data storage will merge and can be queried by an ISO/IEEE standardised SQL:2015 (or so). Because the true spirit of “NoSQL” does not consist in the way data is queried. It consists in the way data is stored. NoSQL is all about data storage. So, sooner or later, you will just create “traditional” tables along with “graph tables” and “hashmap tables” in the same database and join them in single SQL queries without thinking much about today’s hype.

“NoSQL” should be called “SQL with alternative storage models” and queried with pure SQL!