Is this the year you start taking your application performance monitoring more seriously?
With Agile and DevOps practices we’re delivering software to our customers faster than ever, but what happens once it’s in the hands of your users?
Do you have any idea what the customers’ experience is? Are your users experiencing poor performance?
In today’s environment, the customer is king.
There are so many options that if your application is not meeting all their expectations, it’s likely they’ll drop you for a competitor and never look back.
This is where Application Performance Monitoring (APM) (also known as Application Performance Management) can help.
What is Application Performance Monitoring?
An APM solution provides insight into potential performance, bottleneck, and software availability issues with your application you might not otherwise be aware of—until it’s too late.
By being proactive and monitoring all phases of your development pipeline (including production), you can find issues before your customers do.
APM solutions are almost like X-ray vision, allowing you to peer into what would normally be a black box activity and gather insights that can take your application quality to a higher level.
APM enables you to:
- Speed up your development lifecycle
- Perform code audits
- Gain an understanding of which features are really being used by your customers.
- Understand application dependencies
- Uncover performance problems
- Increase your confidence in the software you release
- Monitor the impact of your code changes
- User experience monitoring
If you think about the nature of business in the 21st century, where we hear a lot of talk about the need to innovate, and disruption and blue chip companies going out of business more and more quickly, it’s no longer enough to say ‘we make Band-Aids, and we make the best Band-Aids in the world, and we’ve made them for 100 years, and let’s keep making Band-Aids better and better.’ One of the characteristics of post-industrialism is the need to shift your focus as a company from doing the same thing over and over again to doing new things on a continuous basis. In order to do new things and in order to know whether those new things are relevant, you need some kind of feedback mechanism.—Jeff Sussna, author of Designing Delivery: Rethinking IT in the Digital Service Economy, and my guest on TestTalks Episode 68.
Using application performance monitoring to get the feedback Jeff talked about is essential. Of course, you will need an APM tool for this.
There are many application performance monitoring tools available on the market from trusted vendors like New Relic and AppDynamics, and many of them offer scaled-down, free versions of their products.
If you look at the Gartner Magic Quadrant for Application Performance Monitoring report, all the tools listed are vendor tools like:
What if your team prefers open-source solutions instead?
Here are some of the more popular options according to Github. Regardless of whether you choose a vendor or open- source application for your APM needs, they all should provide you with:
- User experience monitoring
- Business transaction monitoring and track SLAs
- Infrastructure monitoring
- Performance data and response time monitoring
- The ability to do a deep dive into captured data
- Reporting, dash boarding and analytics capabilities
Open Source Application Performance Monitoring Tools List
The most popular APM tool currently in Github is Pinpoint. Designed for large-scale distributed systems written in Java and PHP, Pinpoint is an excellent option for enterprise companies. Pinpoint gives you the ability to analyze the structure of your system and how your components are interconnected using their tracing technology.
Pinpoint’s website highlights the top benefits of their solution:
- The ability to quickly understand your application topology
- Monitor your application in real time
- Get code-level visibility into all your transactions
- APM agent installs without changing any code
Pinpoint currently has 7.8 Github stars.
2. Apache SkyWalking
SkyWalking is an APM solution that is perfect for teams work with Microservices, cloud-native and container-based (Docker, Kubernetes, Mesos) architectures.
According to the Apache website, some of its main features are:
• Service, service instance, and endpoint metrics analysis
• Root cause analysis
• Service topology map analysis
• Service, service instance, and endpoint dependency analysis
• Slow services and endpoints detected
• Performance optimization
• Distributed tracing and context propagation
SkyWalking currently has 5.5k Github stars.
If you have a need a comprehensive monitoring solution for JavaEE applications, check out JavaMelody.
JavaMelody was developed to monitor Java or Java EE applications in QA and production environments.
It’s a tool that allow you to effectively measure and calculate statistics on the real operation of your application depending on its usage.
It also integrates nicely with Jenkins, and can provide you with charts for information like memory, CPU, system load, head dumps and so on.
JavaMelody currently has 1.7k Github stars.
For application performance monitoring of Java server applications, Stagemonitor might be what you’re looking for.
Some of its main features are:
• In-browser widget
• Analyze requests using Kibana dashboard
• Cluster Monitoring—get the metrics of your whole cluster
• Supports creating new plugins
Stagemonitor currently has 1.4k Github stars.
Scouter describes itself as an open-source APM solution similar to New Relic and AppDynamics. Scouter has
- Java agent for Web applications
- Redis, NginX, MongoDB, RabbitMQ and Elasticsearch
It also shows you metrics like:
- User activity
- Service activity including TPS, response time, SQL profiling, etc.
- Resource metrics for CPU, memory, heap
Scouter currently has 1.2k Github stars.
6. App Metrics
• CPU profiling
• Average response times
• CPU, memory, heap monitoring and much more
App Metrics currently has 547 Github stars.
If you’re into Python, Codespeed is a Web application that allows you to monitor and analyze the performance of your code.
Codespeed currently has 439 Github stars.
Even Golang development shops can hop on the APM fun train with GoAppMonitor. GoAppMonitor is a Golang application for monitoring performance data.
According to GitHub, GoAppMonitor is “a library which provides a monitor on your Golang applications. It contains system-level based monitoring and business-level monitoring (custom monitoring). Just add the repository into your apps and register what you want to monitor.”
GoAppMonitor currently has 400 Github stars.
9. App Perf
Although App_Perf is still in the early stages of development, it aims to provide an open-source solution that has similar functionality to vendor tools like NewRelic, AppNeta, SkyLIGHT, etc. App Perf currently has 247 Github stars
APM and Automation Testing
No only does APM data help you to be more proactive with customers, it also can help you create better-automated tests. Having some targeted tests can help you catch issues before a change even gets into production.
“Currently there’s no connection between production and your test. I talk to hundreds of companies every year and almost nobody actually goes on your Google analytics or APM analytics and sees what the coverage is…. but I think [that] if you put an app in production you can learn a lot from that, and you can create tests from that data.”
Make this the Year for APM Awesomeness
With every year that goes by, as more teams adopt CI/CD practices, the need for a better, more effective APM plan is critical to your success. Don’t ignore it. Get some high-level monitoring in place now and start learning from your data.