Welcome!

@CloudExpo Authors: Pat Romanski, Yeshim Deniz, Gopala Krishna Behara, Sridhar Chalasani, Tirumala Khandrika

Blog Feed Post

Challenges with APM 1.0 products

I met several customers in the past few weeks who are evaluating Application Performance Management (APM) solution. They are facing a lot of challenges with their existing investments in old generation of APM solution. In this blog, I will outline some of the shortcomings with APM 1.0 tools that make them unfit for today’s applications.

What is APM 1.0

Customers have been managing application performance since early days of mainframe evolution. However, Application Performance Management as a discipline has gained popularity in the past decade.

 

Let me first introduce what I mean by APM 1.0. The enterprise applications and technologies such as Java have evolved in past two decades. The APM 1.0 tools were invented more than a decade back and they provided great benefits to resolve application issues that were prevalent with the early versions of Java and .NET technologies. However Java/.NET application servers have become mature and do not have those challenges any more. Also enterprise application architecture and technologies have changed drastically and the APM 1.0 tools have not kept up. The following figure shows the evolution of enterprise Java in the past 15 years and when APM 1.0 and APM 2.0 tools have started emerging.

 

image001.png

Following are few challenges with the APM 1.0 tools that you will run into when trying to manage your enterprise applications.

 

Challenge 1: Not enough focus on end-user or visibility for business critical transactions

 

The application owner and the application support team primarily cares about the user experience and service level delivered by their applications. APM 1.0 tools were primarily built to monitor applications from an application infrastructure perspective.

 

These tools lack the capabilities to monitor applications from real user perspective and help you isolate application issues whether it is caused by the network, load balancers, ADNs such as Akamai, or the application, database, etc. Some of these solutions were quick to add some basic end-user monitoring capabilities such as synthetic monitoring. However an application support personnel has to switch between multiple consoles and depend on manual correlation between end-user monitoring and application deep dive monitoring tools.

These tools do not allow you to track a real user request to the line of the code. That means you are blind-sighted when users are impacted and struggle to find what is causing the application failure.

 

Challenge 2: Built for Development and not suitable for production monitoring

APM 1.0 deep-dive monitoring tools were primarily built to diagnose issues during the application development lifecycle. These tools morphed into production deep-dive monitoring tools when the need arose for APM in production environments.  So, These tools were not optimized for production monitoring and hence require a lot of effort to tune for production.

 

First off, the complexities of agent installation and configuration hinder deployment in production environment.  Second, these tools usually require configuration changes every time new application code is rolled out.

 

Most damagingly, they have high overhead on application performance and do not scale beyond 100-150 application servers. This means that most customers use these in a test environment or enable deep-dive monitoring retroactively after an application failure -  assuming the problem will recur.

 

Finally, these tools do not provide operation friendly UIs and because they were originally built for developers.

 

Challenge 3: High Cost of Ownership

 

As I alluded earlier, the old generation APM tools are very complex to configure because these require application knowledge, manual instrumentation and complex agent deployment. Hence expensive consultants are required to deploy and configure and maintain these tools. These tools also have multiple consoles - adding to total cost of ownership. Some customers told me that they spend a lot of time managing these APM tools rather than being able to manage their applications.

 

Conclusion: A Poor fit for today’s applications

These tools were built more than a decade back, and have not evolved much although the application architecture, technologies and methodologies have gone though drastic changes.

 

Many of the customers whom I met were of the opinion that they spend more time managing the APM solution then managing their applications. If you use any of the APM 1.0 tools, and try to manage a modern application, you are likely in the same boat. Here are some customer expectations for a modern  APM solution:

  • It reduces your MTTR by quickly pinpointing business-critical issues with always-on, user-centric, deep application visibility
  • Non-Invasive solution that requires no changes to application code, does not require manual instrumentation and auto-discovers your transactions, frame works, etc
  • It provides Quick Time to Value and Ease of use with a single, integrated APM console
  • Purpose-built for cloud applications

 

APM 1.0 tools certainly cannot satisfy these needs.  In the next blog, I will discuss how an APM 2.0 solution like BMC Application Management addresses the challenges with APM 1.0 products and help you manage applications better thus improving customer satisfaction and resulting in better bottomline.

Read the original blog entry...

More Stories By Debu Panda

Debu Panda is a Director of Product Management at Oracle Corporation. He is lead author of the EJB 3 in Action (Manning Publications) and Middleware Management (Packt). He has more than 20 years of experience in the IT industry and has published numerous articles on enterprise Java technologies and has presented at many conferences. Debu maintains an active blog on enterprise Java at http://debupanda.blogspot.com.

CloudEXPO Stories
Behera Rasananda is a technologist, a leader, a key note speaker has more than 20 years experience in across Government, Financial, Heath Care and Insurance Verticals. Mr. Behera has vast experience in Enterprise Cloud and Big Data solutions and Enterprise Architecture. Currently he works closely for Government Solutions on Enterprise Cloud for Federal Government Agency. Scientist Behera managed and partner with clients to make complete end to end solution and Migration to cloud both private sector and government sector with Cloud Computing and engineering with Architecture, Design and Strategy. This strategy and migration to cloud architecture and enterprise design patterns helps savings organizations of multi-million dollars annually.
Digital Transformation: Preparing Cloud & IoT Security for the Age of Artificial Intelligence. As automation and artificial intelligence (AI) power solution development and delivery, many businesses need to build backend cloud capabilities. Well-poised organizations, marketing smart devices with AI and BlockChain capabilities prepare to refine compliance and regulatory capabilities in 2018. Volumes of health, financial, technical and privacy data, along with tightening compliance requirements by FTC, CUI/DFARS, EU-GDPR and the underlying National Cybersecurity Framework suggest the need for a ground-up re-thinking of security strategies and compliance actions. This session offers actionable advice based on case studies to demonstrate the impact of security and privacy attributes for the cloud-backed IoT and AI ecosystem.
As Enterprise business moves from Monoliths to Microservices, adoption and successful implementations of Microservices become more evident. The goal of Microservices is to improve software delivery speed and increase system safety as scale increases. Documenting hurdles and problems for the use of Microservices will help consultants, architects and specialists to avoid repeating the same mistakes and learn how and when to use (or not use) Microservices at the enterprise level. The circumstance when Microservices is an appropriate solution described in this article and based on the authors' work experiences, best practices available in the Microservices community. However, using Microservices Architecture does not guarantee success to the enterprise.
Today, we have more data to manage than ever. We also have better algorithms that help us access our data faster. Cloud is the driving force behind many of the data warehouse advancements we have enjoyed in recent years. But what are the best practices for storing data in the cloud for machine learning and data science applications?
"Avere Systems deals with data performance optimization in the cloud or on-premise. Even to this day many organizations struggle with what we call the problem of data gravity - 'Where should I put the data?' - because the data dictates ultimately where the jobs are going to run," explained Scott Jeschonek, Director Cloud Solutions at Avere Systems, in this SYS-CON.tv interview at 21st Cloud Expo, held Oct 31 – Nov 2, 2017, at the Santa Clara Convention Center in Santa Clara, CA.