Showing results for 
Search instead for 
Did you mean: 
Create Post

4 Strategies to Improve App Performance

Level 13

By Omar Rafik, SolarWinds Senior Manager, Federal Sales Engineering

Here’s an interesting article by my colleague Brandon Shopp about taking a holistic approach to application monitoring. Applications are important to government customers and Brandon offers some great suggestions.

Optimizing application performance has more to do with the entire application environment than it does with the individual application.

It’s critical for federal IT pros to be able to see the bigger picture by monitoring, mapping, and identifying where there may be larger issues affecting application performance.

Database performance, for example, will absolutely affect applications. System configuration management, dependency mapping and monitoring, and taking a unified approach will also help the federal IT pro down the path of application performance optimization.

Database Analysis

What’s the best way to optimize database performance relative to application performance? Here are several suggestions beyond simple database performance monitoring.

  • Database query analysis – Identify and pinpoint the root cause of slow database queries, then optimize database tables to ensure the optimization of mission-critical applications that rely on them.

  • Indexing Find a tool to help identify and validate index opportunities by monitoring workload usage patterns. Take that even further by investing in a tool that makes recommendations.

  • Flexibility Be sure the agency’s database performance tool works across all environments—on-premises, hybrid, virtualized, and cloud—to ensure a complete picture of the entire application environment.

Systems Configuration Management

Federal IT pros understand the importance of configuration management, especially in terms of establishing and maintaining government IT compliance. A solid systems configuration management tool can help verify system and application patches were installed correctly. For troubleshooting, the tool needs to correlate configuration changes with changes in performance.

Choose a tool that can detect changes on a node (even when offline), and provide customizable reporting, alerting, and performance correlation.

Automated Dependency Mapping

It’s worth investing in a solution that automatically creates custom, logical, contextual maps. With those maps in place, federal IT pros will be able to see application and infrastructure dependencies when troubleshooting a new issue to gain a greater ability to pinpoint the root cause. Once you understand the critical dependencies, you’ll want to be able to drop the key metrics into a chart and share across teams to improve problem resolution.

A Unified Approach

It’s worth reiterating the importance of looking at the whole environment when considering application performance.

As a final recommendation, a “unified approach” means the federal IT team would be best served through a single platform where they can monitor performance across all aspects of the environment at the same time—through a single, unified view. In other words, this type of solution will provide the ability to see application performance, network performance, server performance, storage performance, database performance, etc., all at the same time, as they relate to one another.

Full visibility across all of these elements is absolutely critical to understand where an issue resides and improve uptime and application performance.

Find the full article on Government Technology Insider.

The SolarWinds trademarks, service marks, and logos are the exclusive property of SolarWinds Worldwide, LLC or its affiliates. All other trademarks are the property of their respective owners.

Level 14

Thanks for the article!

Level 12

To improve Application Performance you need to also know what your application/code is doing. Bytecode instrumentation can tell you the average response time, number of executions, hung transactions, errors, etc., down to specific methods in your code. Knowing what the code is doing helps solve problems quicker and gives you other options than throwing more hardware at the problem.


A good network helps too.

Level 15

"It’s worth reiterating the importance of looking at the whole environment when considering application performance."

This is a great comment to take to heart.  I have been part of so many teams that want to simply blame the other areas rather take the holistic view of what the issues are.  A great engineer friend of mine as stated, take an inventory of all possible variables and eliminate them one at a time until you resolve the issue.  That simple thought has been a great impact on the way I review issues. 

Thanks for the posting!

Level 12

I don't think this is what you meant, but it would be nice if IT was a bit more unified. When one person has to call three people on different teams to resolve an issue, and those people may or may not be available then may or may not treat the issue as a priority, it interferes with end user experience and comes across as bad IT.

Level 16

Thanks for the write up.


Thanks for the article!

Level 13

Thanks for the article.

Automatic Dependence Mapping sounds like a dream.  Like it's too good to be true.  I wish we had this.

Level 12

Nice info, any vendors doing this? Would be curious how this is implemented and what software.

Level 11

nice easy read. thank you.

Level 11

Thanks for the post.

Level 20

Often some of the Orion database indexes get really fragmented!

Level 14

The best way for performance to be improved would be to get the software developers (or programmers as we used to be called when I was a programmer) to do their job properly.  Now everything is rushed with impossible deadlines and limited testing.  That just leads to poorly performing applications. 

I wrote some software for a company in 1988.  As Y2K approached the finally tracked me down and were very happy to learn that I had catered for it in the code.  Pity.  I could have earned quite a lot of money.  Probably should have lied and agreed to 'fix' it for a large fee but I'm too honest.

Level 13

Let's stop pointing fingers.  Thanks for the info.

Level 16

I would love to see automated dependency mapping working.

Level 13

Yep yep yep.  Unified team approach to the problem is what you want - unfortunately hard to achieve.  Most people (and vendors) are looking to finger point instead of rolling up their sleeves and working together to fix the problem. Just own it and fix it folks.