The world of the Database Administrator (DBA) has changed since the technology went mainstream in the 1970s, writes Bruce Momjian, co-founder of the PostgreSQL Global Development Group. Over the last 40 years three or four vendors have sought to dominate the market and have invested heavily in building the “Rolls Royce” of databases for mission-critical on premises applications.  When the Oracle database marked its 30th birthday in 2007, it felt almost like a cult gathering as everyone in the Moscone Center gave Larry Ellison a standing ovation as he arrived for his keynote.

It was an understandable reaction, because the DBAs in the audience had also invested heavily in becoming Oracle specialists.  That 30th Birthday celebration was symbolic of the unspoken contract between DBAs and Oracle at that time.  Both were heavily dependent on one another.  Fast forward to today, though, and the relationship is changing.  More and more DBAs are questioning their on-going reliance on Oracle.  When I meet Oracle customers, I ask them one deliberately provocative question:  assuming that you are happy with Oracle today, will you still be happy in five years time?

DBA
Bruce Momjian, co-founder of the PostgreSQL Global Development Group

Given how close the relationship between DBAs and Oracle has been historically, it is interesting to note how individuals respond.  Initially there is some amusement or dismissive snorts of “Of course I will!”

But when we begin to unpack how market dynamics have changed and consider the new opportunities that exist because of open source, the mood changes.

More fundamentally as we explore what these DBAs need from their database platforms it prompts reflection and acceptance that while Oracle has been the defacto database standard on premises it may not be the best option moving forward, particularly as organisations seek greater flexibility and agility from their database infrastructure.  By the end of these sessions we often have a very energised audience wanting to go out and explore the potential of open source-based databases such as PostgreSQL.

Today’s Reality: Oracle and the DBA Community Drifting Apart?

Given that the database is so critical to Oracle’s Cloud strategy, you would think Oracle would do everything possible to maintain its traditionally strong bond with the DBA community.  However, as with any relationship, circumstances change and sometimes those involved drift apart. So what has caused this relationship to break down?

F1 On-Premises Database

Oracle has developed an on premises database that has so much functionality it could be described as the Formula One equivalent of a database. I would stick my neck out and say that 95% of the configuration files for the vast majority of databases have not been changed since deployment.  That suggests DBAs do not need all of the functionality Oracle offers.  Gartner has said that open source databases could handle 95% of all current database requirements, so it should be no surprise that developers are turning to open source alternatives for faster, more agile databases that deliver what they need, not what the vendor thinks they need.

Barebones Cloud Database

While DBAs are used to feature-rich on premises Oracle databases the same cannot be said for the Oracle Cloud. Gartner has suggested that Oracle Cloud is a minimal viable product and that is hardly surprising, because Larry Ellison has admitted that Generation 2 of the Oracle Cloud has required a complete re-architecture of its technology.  This should serve as a warning that DBAs should not expect like-for-like equivalency in the Cloud!

The Autonomous Database

Oracle sales executives position it as a way to free up DBAs to do more; however, their bosses are more interested in appealing to business decision makers, explicitly saying it will be an opportunity to cut costs…the most expensive of which is the DBA running the database! Despite this blunt threat the reality is that autonomous technologies will, in the long-term, become a feature of database technology.  This impending change is encouraging DBAs to re-evaluate their database strategies; this is not necessarily a reevaluation in Oracle’s favour.

The Cloud Changes Market Dynamics

In today’s digital world, where agility and cost are the main drivers, the business case for Oracle is becoming harder to justify.  As automotive manufacturers talk about a future with mobility as a service rather than consumers owning cars, similarly DBAs are being told they don’t need to “own” databases on premises.  Rather they should switch to Cloud-based services, but the danger is that while these “all you can eat” contracts seem appealing, customers can quickly rack up huge bills for databases that are not fit for the challenges they face. Developers want to spin up instances far more rapidly and at less cost than traditional on-premise databases. The Cloud gives them the ability to move away from the likes of Oracle to cloud-native options that offer greater flexibility and are more fit for their specific purposes.

The Alternative: A Time for Pragmatic Reflection

I realise that for Oracle diehards what I am saying is nothing short of heresy, but there are many pragmatic DBAs out there who recognise that a schism is happening.  The combination of the on-going fallout from the financial crisis (which has suppressed IT budgets for many years) and the urgency of responding to the digital era is forcing cool-headed, pragmatic reflection.

Take, for instance, Oracle Real Application Clusters (RAC).  It was originally developed as AWS was only getting off the ground and Larry thought the future was Grid Computing (he famously called Cloud “complete gibberish” back then!)  It took many years to perfect, but it was positioned as the solution to the highly performant, reliable databases that everyone predicted we would need for Internet 2.0.

Today, we hear commentators talking about data lakes, which in theory should be ideally suited to RAC.  In reality, RAC is highly complex, costly and not scalable enough for such tasks.  It is why Oracle has struggled to gain traction with the internet unicorns that are dominating business today.  They have turned to a wide variety of open source-based alternatives that can provide the more immediate and appropriate functionality needed to scale their businesses.

Even the well-established internet giants – those who you would expect would need the full might of Oracle database functionality – are seeking to break away.  Last year there were rumours about Salesforce looking to move off Oracle and more recently, Amazon Web Services (AWS) has been frank about its desire to move off Oracle. Of course, this makes for great media headlines, but the underlying message is clear – Oracle is expensive, the functionality isn’t ideally suited to the digital era and there are many credible alternatives available. 

The Future: Do You Have a Five-Year Database Plan?

We shouldn’t be surprised that many DBAs view Oracle as a legacy database platform that is not fit for the demands of digital business.  Increasingly, I expect we will see DBAs answering my original question by moving away from Oracle, because it cannot offer the level of innovation they will require in the future.  And for those who are still unsure or evaluating their options then I would recommend some simple tests:

  • Is there a practical reason for choosing an Oracle database?
  • Does this application require specific functionality in Oracle?
  • How much do you want to spend?
  • How much configuration do you need to do?
  • If you need scalability, can the Oracle Cloud deliver it?

Whether you are considering alternatives to Oracle or believe Oracle is the right option, the most important point is that you must start planning as early as possible.

There are many factors to consider from the technical requirements to the licensing implications of your database choice – both at time of implementation and into the future. Above all it is important to choose the database that fits your business needs.  If the direction for the IT department is specific and clear, then it may be sensible to use a closed source database such as Oracle.  However, in this day and age it is rare that a business has such certainty.  If your business is demanding greater flexibility from IT, because the strategic direction is fluid, then it makes sense to consider open source-based alternatives.  This will give DBAs the opportunity to adjust the database strategy as the direction of travel for the business changes, allowing for the exploration of different markets and customer demand.

Bruce Momjian is co-founder and core team member of the PostgreSQL Global Development Group, and has worked on PostgreSQL since 1996. He has been employed by EnterpriseDB, the database platform company, since 2006