In de border area between Database Administrator (DBA)  and the application or its developer.

Open Source

Closed Source

Database Optimisation Specialist

As a Database Optimisation Specialist, I focus on optimising the database as a whole and everything that influences its behaviour, among which:

  • Functional use of the database
  • Technical use of the database
  • Applications that use the database
  • Table structure
  • Datatype use
  • Relational model
  • Setup of the database cluster /  instance
  • Setup of the underlying components like operating system, storage, memory and network.

Database use

Important part in this is to verify the specific queries that the database has to handle. These often are generated by other applications and are not always efficient enough. By checking use and behaviour of the database in a combined force with application managers or application developers, usually a lot can be optimised by some simple steps. Often a few adjustments in the query approach is enough, while sometimes it requires an improvement of the datamodel.

Maintenance

Another common issue with databases is found in lack of maintenance. Once started small with a little amount of data and little use, then grew successfully without thinking about the consequences for the database. For a properly setup database receiving proper queries, there’s hardly any impact when datavolume and/or usage grows. Each database however does require regular maintenance. Always, just like the best car ever requires regular maintenance.

System

Any database is as fast as its slowest component. Therefore it’s important to make sure underlying components like operating system, harddisk, memory and network really align with the specific database use. Most Database Administrators (DBA’s) maintain many databases, hundreds or even thousands. For a DBA, it’s impossible to know, understand and remember all details for each database. Therefore they usually stick to generics for support and optimisation.

Many application developers or application managers however, lack the knowledge to provide the DBA with proper information. By result the database often is tuned pretty well on generic level, but could be tuned much better in the details. The datamodel and the using the right query often remain neglected, because the DBA doesn’t have the correct detailed information about the specific use of that specific database.

The Database Developer acts as a middleman because this role understands both sides well enough to fill in the “missing link”.  

The numbers tell the tale

One very underestimated subject in Database Optimisation is the world of measuring, building up history of metrics and visualising these metrics. As I have a background in Business Intelligence, I know the power of measuring and analysing these metrics over time. Although most (if not all) database engine have views showing useful metrics, hardly any organisation uses them to the full. Oh sure, some developers or database specialists know of them and sometimes use them, but hardly any organisation captures them to analyse them over time.

But when you start doing so, the real power arises. Sure, it takes time to set something up to collect and visualise them, but once there, it immediately shows the real pain points. When those are taken away, the next pain points pop up. Gradually eliminating all pain points, ending up with a system that runs as smooth as possible.

And guess what? If your system runs smooth, it’s not wasting resources and thus saves large figures of money because you can downsize your virtual server. Or don’t have to upsize because the system grows. The latter immediately argues why, once created, the metrics collector needs to stay and be monitored regularly. Being the only way to make sure your system stays running smooth on the long run.

I’ve set up such systems several times now, using Open Source solutions which keeps costs low. It does take time to set this up, it certainly is an investment, but it pays off quite fast!

Difference between Database Optimisation Specialist and Database Administrator

The Database Optimisation Specialist mainly focusses on the specifics of practical use of the databases, how the database behaves internally and how the practical use can be aligned with this internal behaviour. A DBA focusses more on generic setup, generic maintenance and security (login). 

Additionally, many Database Administrators (DBAs) know one or two database engines very well, but not more. As I work with developers and application operators, I move along with their needs. Meanwhile, over time during my career, I have worked with several database engines. Due to this, I am fully aware of the similarities and differences. This makes that when I come across a database engine that I don’t know that well yet, I pretty quick get to the bottom of its specifics.

By result I now have in-depth knowledge of many databases, in some areas even better than the specialised DBA. Then again, a DBA has its own work area of which he/she knows far more than I do.