Chuck McDevitt: An Obituary

Teradata Storage Rack
Teradata Storage Rack (Photo credit: pchow98)

My friend and a major contributor to the art of database architecture, Chuck McDevitt, died last week. Five years ago Chuck was diagnosed with an advanced cancer and given six months to live. He fought and endured and worked for most of those five years, teaching us all a little about how we might live our lives.

Chuck’s contributions to database architecture are not so well known. He was employee number fourteen at Teradata and developed, with Dan Holle, the Teradata version of SQL. Chuck invented several foundational parts of any parallel database system. He left Teradata and went to Cogit, a start-up that developed a very early parallel data mining tool. From there Chuck went to Ab Inito as a senior architect, and from there he went to Greenplum where he was the Chief Architect.

At Greenplum Chuck was the brains behind the development of their parallel version of Postgres. It is significant to note that Chuck’s architectural insights led to an extensible and powerful implementation that far exceeded the efforts of others trying to accomplish the same result from a Postgres starting point: Aster Data, Netezza, DataAllegro, and the Postgres community.

To convey Chuck’s contribution to Teradata let me tell a story.

After E.F. Codd published his ground-breaking papers on relational theory two research projects began to develop relational databases. The UC Berkeley project developed a query language that was called QUEL which was followed by an IBM Research project and a query language called Sequel. As Teradata entered the scene they developed a third query language that was called TEQUEL, selecting the best features of each. IBM then delivered DB2 and SQL/DS, and the Oracle Database appeared, all based on Sequel which was by then called SQL. QUEL was forgotten as SQL dominated the market leading to it becoming an ANSI standard in 1986.

One Friday in 1985, the Teradata sales management convened a meeting to request that Teradata support SQL in order to effectively compete in the emerging RDBMS market. They were told by Product Management that it would take three years to develop SQL support… that it was out of the question. Further, the Teradata CTO argued that TEQUEL was superior to SQL and that this was an advantage in any case.

Dan Holle heard about the meeting, called Chuck in and, convinced that the Teradata management team was headed in the wrong direction, started to work. What the technical Execs must have known, but clearly did not appreciate, was that Dan and Chuck had developed the Teradata database using what was at the time, a very advanced concept… compiling the TEQUEL query language into an intermediate language and then processing that. Working non-stop through the weekend Chuck developed an SQL parser that would generate the proper intermediate code for consumption by the optimizer and execution engine… and on Monday morning he demonstrated a functioning SQL version of Teradata.

It is fair to say that, but for Chuck McDevitt, Teradata would not exist… and it is likely that Greenplum would not exist.

As I repeatedly suggest in this blog… architecture counts… and you should all know that the database community lost a great architect last week. Chuck will be missed.

30+ Year Old Database Architecture: DB2, Oracle, Postgres, Teradata, Sybase, and More…

As you look at the enterprise RDBMS marketplace today you will find something shocking… almost every product in the market is built based on designs and concepts that are over thirty years old. IBM’s System R grew into DB2 and influenced Oracle before 1980. Ingres, developed before 1980, became Postgres which became Netezza and Greenplum and more. Teradata was a fresh start… around 1980.

This is not a bad thing in its own right… but imagine the hardware architectures these systems were designed and optimized for. Maybe DB2 was built for a multi-core mainframe… maybe Oracle too… maybe. Memory was tiny… so memory management was important and memory was used sparingly. Data sizes were tiny. Consider the fact that Teradata named the company based on the belief that someday way beyond the planning horizon some customers might get to a terabyte of data.

The reality is that these old designs are inefficient. They have hacked the old code to continuously extend their products. I mean this as a compliment. It is not trivial engineering to find tweaks and tack-ons that make old code work on new hardware architectures. Teradata and Netezza and Greenplum designed ways to use multiple address spaces to take advantage of multiple cores. Oracle tacked-on a shared-nothing I/O subsystem to a shared-everything architecture to stretch.

But these hacks are not efficient.

Yale is working on some new-new stuff (see here). HANA is based on a completely different design (see here). The NoSQL vendors have bent the ACID-tested rules, if not always the fundamental approaches.

I can’t help but believe that in one of these new approaches is a path forward.

If you would like to read some history of the start here is a cool link.

%d bloggers like this: