Here is an interesting article from my colleague Joe Kim, in which he explores database performance management.
Databases are complex, multifaceted, and vital to the health of every agency. They are the heart of every data center and arguably one of the most important components of an agency’s technology infrastructure, whether on-premises, in the cloud, or within a hybrid IT environment
For these reasons, optimizing database performance is critical to enabling an optimized data center. There are five things a federal IT manager can do to meet this goal:
- Ensure that databases are healthy
- Gain visibility into data and metrics
- Put data and metrics in context
- Track optimization plans
- Create and maintain a performance baseline
Let’s look at each of these steps individually to understand how they all fit together:
Ensure Database Health
When it comes to databases, health and performance are two different things; databases must be healthy before they can be optimized. Signifiers of database health include things like CPU utilization, I/O statistics, and memory pressure. Collectively, these metrics can indicate whether a database can perform well.
The next step is to begin the performance optimization process, ensuring that queries can execute quickly and throughput can be maximized. This starts with gaining full visibility into the data and metrics needed to assess database performance
. For example, the ability to drill down into granular metrics like resource contention and a database’s workload is key in helping to identify and mitigate the root cause of a performance problem.
Put Data in Context
Agencies should be sure that data delivered by your monitoring tool
is structured and presented in a way that gives you and your team the real insights necessary to fix problems and optimize performance. Specifically, the data should help quickly identify and resolve the root cause of performance issues, and not lead you down a “rat hole” of unnecessary, second-level research and analysis.
Track Optimization Plans
There are things your team will do to test and optimize performance, including running optimization queries, for example. Make sure all queries and tests are tracked, and that results are carefully correlated with the tests being performed.
Create and Maintain a Performance Baseline
It’s nearly impossible to tell when a database is underperforming if you don’t have a daily baseline “normal” to measure against. The best approach is to implement a comprehensive series of management and monitoring tools.
Make sure you also have the ability to view down, up, and across; it should let you drill down into the database, across database technologies, and across deployment methods (including cloud
). And, finally, make sure the tools you choose allow you to establish a historical record of performance metrics.
All this information, coupled with the ability to create a baseline, will help ensure that your IT teams have the tools they need to optimize the health and performance of their database.
Find the full article on Federal Technology Insider