APMdigest asked experts from across the industry – including consultants, analysts and the leading vendors – for recommendations on the best way to ensure application performance in the hybrid cloud. Part 2 covers BTM, NPM and ITOA.

Start with 28 Ways to Ensure Application Performance in the Hybrid Cloud - Part 1

5. BUSINESS TRANSACTION MONITORING

The secret to ensuring app performance in hybrid cloud or other complex environments is to focus on the business transaction. Such transactions connect the end-user experience to the end-to-end performance of all relevant elements of the architecture as the transaction traverses whatever cloud environments and on-premise systems are necessary to achieve the business goal. In contrast, if you think of a hybrid cloud environment as separate environments that you've connected together, application performance can be difficult to understand, let alone optimize.
Jason Bloomberg
President, Intellyx

The foundation of any APM tool is the collection and analysis of transactional data captured at the application layer. In hybrid cloud environments, users must ensure that the APM deployment is able to instrument the nodes which support application traffic flows, regardless of where that code executes, in order to capture a complete view of the code execution path(s) and therefore any performance bottlenecks. Failure to do so ensures that performance engineers will have application performance blind spots – they will easily find the problems in the areas they can see, but inevitably, performance issues in the blind spots will go undetected. Having complete visibility across the stack enables organizations to proactively address performance issues before they impact the business and ensure a consistent, high quality end user experience.
John Maxwell
VP, Product Management, Dell Software

The main thing is to ensure that the flow of activity between the private and the public infrastructure and applications is maintained, as if both infrastructures reside in the same network. The application's end users should not be affected in any way by the hybrid approach. Transaction tracing, which must be completely transparent and agnostic to the underlying operating systems and the hardware (physical or logical) it utilizes, is the key to assuring performance. By constantly monitoring the flow from public-private-public-private etc. components, we can assure the overall solution is performing as designed.
Zvika Meiseles
CTO, Correlsense

6. AUTOSCALING

It isn't enough to just have end-to-end visibility into a business transaction deployed across a distributed hybrid cloud environment. The ability to auto-scale the cloud applications using health rules and policies based on a combination of application and infrastructure metrics is also important. For example, If the average response time deteriorates, auto-scaling an e-commerce application deployed in an AWS infrastructure will come in handy. This not only helps deliver exceptional end-user experience, but also minimizes the operational cost by preventing over provisioning of the infrastructure.
Anand Akela
Director, Product Marketing, AppDynamics

7. NETWORK PERFORMANCE MONITORING (NPM)

To network infrastructure teams, I recommend that they evaluate the tools they use for network performance management. They need to determine whether their existing tools can provide them with end-to-end visibility across internal, private cloud infrastructure and external, public cloud infrastructure, as well as visibility into the network connectivity that links the two.
Shamus McGillicuddy
Senior Analyst, Network Management, Enterprise Management Associates (EMA)

With hybrid cloud, application performance suddenly becomes dependent on network and Internet performance – something that most application developers do not consider when they design their apps. As such, anyone moving applications to hybrid cloud needs to have a tool for providing end-to-end network performance from the end-user through the WAN to the application.
Damian Roskill
CMO, AppNeta

8. NPM + END USER EXPERIENCE MONITORING

The tools you've been using to monitor applications in your data center can't see far enough into SaaS-based applications. Since the network is the resource that connects your user and your applications, wherever they may be hosted, network performance monitoring provides you the needed visibility as you transition to the cloud. Pick a platform with a focus on end-user experience; a combination of network performance and end-user experience is an effective monitoring technology to ensure application performance in the hybrid IT environment.
Ulrica de Fort-Menares
VP of Product Strategy, LiveAction

9. ITOA

IT Operations Analytics technology detecting the latest state of the hybrid cloud environments and tracking history of the introduced changes is essential to creating visibility into an otherwise black box of the cloud.
Sasha Gilenson
CEO, Evolven

Read Sasha Gilenson's blog: ITOA - Essential for Hybrid Cloud

Hybrid cloud environments add another dimension of variability to the already diverse set of transaction types present in today's complex IT infrastructures. When attempting to diagnose application performance issues, it can be especially challenging to assess the impact caused by the various network paths present in hybrid cloud environments. To ensure optimal performance, IT Operations and DevOps teams must turn to advanced behavioral analytics to detect anomalies in various application performance metrics, and then statistically link them to other anomalies that may be present in various network and cloud-based performance metrics. Traditional network monitoring that looks only at aggregate bandwidth levels is insufficient to correlate with specific application level problems. Behavioral analytics employs automated machine learning algorithms to create baseline models of normal behaviors of all transaction types within the hybrid cloud network, detect unusual behaviors, and link them to other application anomalies, enabling speedy root cause identification of performance issues.
Mike Paquette
VP of Products, Prelert

10. LOG ANALYTICS

You need more than one tool and approach, but the one I recommend is a central log analytics solution. Hybrid cloud environments are complex systems with a variety of distributed components that make up your app. Each of these components (hopefully) writes log data, and only if you collect all those logs in one central place and have the ability to analyze it in meaningful ways you will be able to identify and eliminate performance bottlenecks. Consolidated and aggregated log files from all your components will give you a cohesive picture of that entire modular system. Do not allow blind spots!
Jon Gifford
Founder and Chief Search Officer, Loggly

Read 28 Ways to Ensure Application Performance in the Hybrid Cloud - Part 3, covering monitoring and visibility.