As a consultant at Oracle 25 years ago, I was well aware of Ingres, the relational database developed as a student project at UC Berkeley in the ‘70s. In the early to mid-‘80s Oracle and Ingres were the two most prominent vendors in the then nascent relational database market. In fact, many experts at the time preferred Ingres with its QUEL query language to Oracle and SQL. Guess who won that showdown?
Ingres survived over the years under various owners including Computer Associates. They were ultimately spun-off from CA as an open source company. Unsympathetic critics decried Ingres' post-CA business model as “orphan source.”
OpenBI established a relationship with Ingres about five years ago. It seemed a natural. Ingres was an aggressive open source technology company and OpenBI was a BI professional services firm focused on OS software solutions.
Turns out the synergies were not as fruitful as first thought. Ingres had no special history in BI and in fact lagged behind MySQL and Postgres as an easy-to-use relational database for the OSBI market. Indeed, OpenBI found Ingres and its tools frustratingly buggy and much more difficult to get up and running with our chosen BI platforms than either MySQL or Postgres. So with no legacy Ingres customers to call on, OpenBI didn't see an ongoing business proposition with Ingres and abandoned the relationship two years ago.
It appears quite a lot has changed at Ingres since the divorce. The landing page of the Actian website boldly greets surfers with “Ingres Becomes Actian to Take Action on Big Data”. Hmm, Big Data – Ya think?
The new Actian product portfolio includes:
- Action Apps – lightweight, consumer-style applications that automate business actions triggered by real-time changes in data.
- Cloud Action Platform – the industry’s first development environment for Action Apps.
- Vectorwise – the world’s fastest analytical database – “record-breaking” action engine for big data.
- Legacy Ingres database and related tools.
Even though I don't know enough to separate current product reality from marketing bravado, I applaud Actian's attempts to re-brand with a new strategy surrounding big data, analytics, cloud processing and decision management. May Ingres rest in peace.
Intrigued with what I'd read on the website, I decided to download the 30-day Windows trial version of Vectorwise and put it through a few exercises during the down times of the long Thanksgiving weekend.
The installation went smoothly, though I banged my head against the wall attempting to get the visual import GUI to work. Recalling frustrations from the past, I was close to abandoning the effort but decided instead to dig into the User Guide. There I found write-ups on command line programs to accomplish pretty much everything I wanted to do with the Vectorwise server. I could create databases, build schemas, import data, run SQL batches, backup databases, etc. – and drive all from shell scripts. I was back in 1985 DBA heaven. So what if my partners poke fun at my antediluvian methods?
The results I've gathered so far for my admittedly non-strenuous tests are nonetheless encouraging. My first experiment was loading a 600,000 row fact table with nine small star lookups from csv files. The big table load takes two seconds and all the queries I've attempted, even ones with a six table join, complete in a second or two.
The second test involved loading a 10 million+ row, four attribute, csv stock performance data set along with a 3000 record lookup. The big table imports in 7 seconds. Group-by queries that join to the lookup complete in under three seconds.
I've not as yet, however, been successful connecting through the Vectorwise's ODBC and JDBC drivers. Robust ODBC/JDBC support is of course critical for linking the database server to front-end applications and development tools. Programs of special personal interest include the R statistical platform, the Pentaho BI suite, especially PDI, and my beloved Ruby. If memory serves, OpenBI struggled with the Ingres JDBC drivers several years ago. I'll reach out to Actian for support and report my findings in a future blog.
So far, I've found my investment in Vectorwise to be productive. Though I wouldn't dignify the efforts as a true evaluation, I'm impressed with the performance I've seen on my small “proof of concept.” In the near future, I'd like to determine how the platform stands up to hundreds of gigabytes/terabytes of data and scores of simultaneous users. I'd also want to see how it rates as an analytic data store for development with OS platforms Pentaho and Jaspersoft.
Even with an incomplete grade, I've seen enough positives to continue evaluating Vectorwise and to engage Actian for intermediate-sized BI/analytics opportunities. I could certainly envision Vectorwise assuming a position in the mid-sized analytical database usual suspects list that for OpenBI includes Vertica, LucidDB, Infobright, Infinidb and Calpont. Perhaps with continued good behavior the orphan can find a home!