Regardless of your infrastructure whether you are running in the cloud or on-premise, there will always be a need to “tune your Database”. I use the double quotes because this phrase seems to be a catch all statement used for a myriad of underlying issues. A seemingly obvious question; Is it really the Database that is under-performing or is it in fact some use case that is being forced to take the long way around?
The short-term response normally consists of adding a few indexes here and there, but this only seems to befit specific use cases and maybe improves the efficiency slightly.
Nine times out of ten the medium-term response is to upgrade to the next level of infrastructure, be it the next generation of hardware or the next type of Virtual Machine or Instance. This is in the hope that more CPU Power or memory will improve or mask the issue. A solution that typically only lasts for a few months.
Well, you may have successfully migrated or upgraded your system from A to B, but have things improved by the magnitude you had envisioned. Or have some processes even degraded? In all honesty, we see these band-aid solutions and outcomes all the time.
So, what is the long-term approach? Let’s take a step back and review the common situations enterprises find themselves.
Your datasets have grown from Gigabytes to Terabytes in just a few years or even months
You have heavily customised your out-of-the-box applications
No one is looking at the distribution of the data
Could the performance problems be due to how the data is being accessed and retrieved? Maybe the set of queries that process the monthly billing run or produce the scheduled reports are now two years old and are no longer appropriate for the volume, type or method by which the data is stored. Other than re-writing the application code what else can be done?
One of our first activities is to focus on the Data Management. It seems to me we are constantly being told how important data is for business but does your organisation have a clearly defined:
Data Architecture
Data Design
Data Partitioning Strategy
Data Archival Strategy
Data Clustering Strategy
These Data Management fundamentals should form the basis of every successful operation, especially now with the expectation of Big Data and Machine Learning. It is imperative that we get the basics right, because without good foundations how do we expect to reach the next level!
Consult with Fusion Professional to discover how our best practices can ensure your enterprise data management is not holding you back.
Comments