Network-based Application Performance Monitoring solution measures delays in network and application for all transactions of all users. If any problem occurs, it immediately reports and alerts the administrator and provides all necessary data to point out the cause of the performance issues. It is often compared to traditional APM solution, so let's see where the limits of such agent-less solution are?
Application Performance Monitoring (APM) detects and reports performance issues to ensure smooth operations, satisfied customers and internal users working efficiently. Most of us experienced a situation of availability issues or latency of web service. It is often difficult to find out the root cause quickly. Network department finger points at the application department or an external supplier of the application and vice-versa. But where is the truth?
APM tools are here for such situations. May be, you consider APM as robust and expensive solution. And I have to agree, that the traditional APM approach is both robust and expensive. However, there is a cost-effective approach with no agents, relying on network data only. Of course, there are features requiring agents and network based solution cannot help in some use-cases. Let's compare agent and agent-less approach in Application Performance Monitoring.
APM Approach Comparison
We can see in the chart above, main differences are in the source code execution visibility and tracking transactions through application chain. These features need agents and network-based APM cannot deliver such a visibility. On the other hand, network-based approach is completely independent on the operating system and applications. It represents passive monitoring, with no impact on infrastructure and applications so the deployment is a matter of minutes.
Use Cases For Network-based APM
Let's see typical use-cases, where Flowmon APM representing agentless apprach is an efficient option for measuring of application performance.
Online Service Provider
Customers provide on-line services (e.g. Internet Banking Application, Online Gaming, E-shop, etc) and every downtime of an application causes loss of customer satisfaction, damage of reputation and consequent financial losses.
- Flowmon APM permanently measures performance indicators of each transaction and reports fulfilled and missed SLAs.
- In case of decreased performance, Flowmon APM helps in fast location of problem and root-cause analysis.
- When releasing a new version of an application, Flowmon APM provides data for evaluating how the new version performs compared to the previous one.
Business-critical Application
A customer’s business depends on internal systems consisting of applications and database servers (e.g. ERP, CRM, Internal Information System, etc). Every minute of system downtime results in employees not being able to do their jobs and the business stops.
- After deploying Flowmon APM at the customer’s side, each transaction is permanently measured and problems are reported before they affect the business.
- Intranet system and its parts (modules) are monitored from the network, application server and database server perspectives.
- Flowmon APM reports performance issues for specific parts of an application and points at where the delay comes from (e.g.network, application server or database server).
One the othere hand, there are use-cases which require agents and network-based approach is not an option:
- Monitoring of execution at source code level
- Visibility of processes at operating system level
- Active application checks using synthetic users
Summary
Network-based APM is a cost-effective alternative for customers requiring an easy-to-deploy solution based on network traffic capturing and analysis. Distinguishing between network delay, application and database delay when monitoring user experience is the main benefit. It is the perfect choice for companies running HTTP/HTTPS applications and databases with lack of application performance visibility. Explore Flowmon APM, agent-less application performance monitoing solution.