Click here to close now.

Welcome!

CloudExpo® Blog Authors: Pat Romanski, Elizabeth White, Lori MacVittie, Carmen Gonzalez, Roger Strukhoff

Related Topics: CloudExpo® Blog, @MicroservicesE Blog, @ContainersExpo Blog

CloudExpo® Blog: Article

Challenges and Best Practices for Load Testing with the Cloud

Part 2: The key is to think of cloud testing as the delivery mechanism

In Part 1 I described how the cloud is revolutionizing load testing and the advantages it provides to ensure that your web applications perform well in production. We also looked at what capabilities you should seek out when selecting a load testing solution.

In Part 2, I will offer the limitations of a test strategy that relies solely on cloud-based testing, highlighting the need for a complementary internal load testing solution. I will also discuss several best practices for load testing in the cloud. Understanding how to apply the right tools and practices to make the most of the cloud is fundamental to cloud-based testing and vital to ultimately going live with total peace of mind.

The advantages to load testing with the cloud are clear, but internal testing still has its place in the overall test plan, particularly when testing from outside the firewall is not feasible. Internal testing also helps you to isolate effects that are due to your own application or infrastructure from those that are outside your firewall and potentially beyond your control.

The advantages to load testing with the cloud are clear, but internal testing still has its place in the overall test plan, particularly when testing from outside the firewall is not feasible. Internal testing also helps you to isolate effects that are due to your own application or infrastructure from those that are outside your firewall and potentially beyond your control.

Isolating Root Cause
When load testing uncovers a problem, the next step is identifying which layer in the delivery chain is causing the problem. You can use monitors to check performance metrics such has hits, average response time per request, and average bandwidth for each layer or piece of equipment in the chain. You can also employ application performance management (APM) solutions to identify bottlenecks in the code. These techniques work reasonably well when there is a single source for the performance slow-down.

When there are multiple problems, both inside and outside the firewall, it can be difficult to sort out the root causes because the symptoms are often commingled, making them difficult to isolate and resolve individually. For this reason it's important to have a cloud load testing solution that you can also apply within the firewall. You can then separate the problems that exist within the firewall from those caused by layers outside it. The ability to test a subset of the delivery chain in this way makes it much easier to find the root causes of performance problems.

Reproducing Tests
Often, you need to precisely measure the effect of changes made to the application code or settings. For example, you may need to determine the performance improvement that results from resolving a specific defect or evaluate performance for a range of cache sizes and other settings.

With cloud load testing, such precise measurements are difficult because of variations in Internet traffic and bandwidth availability at the data center level. Such variations can make it almost impossible to duplicate conditions from day to day or even within the same day.

As with isolating root causes, this situation also highlights the need for internal testing, in which you can better manage the conditions of the test, stabilize the testing environment, and take precise measurements to get more reliable performance metrics for comparison.

Conducting Small Scale Tests
Not all load testing requires hundreds of load generators. Even applications that anticipate many thousands of concurrent users are initially tested with a small population. These smaller scale tests require only a few machines may be easier and less expensive to conduct internally if the machines have already been purchased and are available for use. These tests can be carried out within the firewall to conduct tests that don't require a heavy load or the full delivery chain. To keep cloud expenses down, use available real machines when they can provide the information you need, and employ load testing with the cloud when necessary for larger scale, more realistic tests.

Testing Inside the Firewall
Of course, some testing use cases preclude the use of the cloud. If you're developing an enterprise web application that was not designed to be accessed from the Internet, then it's not a good idea to expose it outside the firewall solely for the purpose of load testing with the cloud. Likewise, if your pre-production environment is not set up to be accessed from the Internet, you'll want to have an internal testing solution that can be used within the firewall. Ideally, you want to use the same load testing solution for both internal testing and testing with the cloud, so that the scripts you developed for internal pre-production testing can be reused in production cloud-based testing. Using different tools for internal and cloud testing would not only require a rewrite of the scripts, it would also increase licensing and training costs.

Best Practices
The following best practices can help you maximize the advantages - and minimize the challenges - of load testing with the cloud.

Employ a Two-Stage Process
A two-stage process for load testing enables engineers to employ internal and cloud testing in the situations for which they are most effective and appropriate. In the first stage of the process you conduct internal tests with a medium load to quickly identify and resolve preliminary performance issues. Then increase the load incrementally with one or more load generators in the test lab. After the performance has been validated internally, proceed to the second stage, cloud-based load testing, for large scale tests that validate the entire delivery chain of the application.

This hybrid approach addresses the key challenges facing organizations that attempt to rely on testing from the cloud only:

  • It enables teams to isolate problems. The source of any performance issue identified in the first stage is clearly within the firewall (because no other systems are involved in the test). It's easier to pinpoint and fix internal problems when they are not being compounded by other issues that originate outside the firewall.
  • It enables earlier testing. With the two-stage process, you don't have to wait for the application to be deployed and accessible from the Internet to test it. You can test internally earlier in the application lifecycle when defects are easier and less expensive to fix.
  • It enables reproducible tests. With internal testing you have much more control over the environment, so you can precisely measure the effect of code or configuration changes on application performance.
  • It provides a better understanding of each major area of the delivery chain. You can compare the results of the same test scenario run internally and from the cloud to get a clearer picture of how the application server and network infrastructure contribute to overall response times.
  • It lowers costs. Cloud testing is based on a pay-per-use model. When you can test internally on hardware you already have, you can reduce the amount of testing that you need to perform from the cloud and cut costs.

Use Different Cloud Providers
There are several advantages to using multiple cloud providers. First, it helps you test from more geographical regions, which provides more realistic results that capture the effects of various third-party servers and content delivery networks. Second, it's more scalable. For exceptionally large scale tests, you can engage multiple providers simultaneously to bypass limitations that a single provider may place on bandwidth or the number of machines in use. Third, it enables you to detect potential network issues at the cloud provider level. If test results from virtually all providers show acceptable performance, but you're seeing significantly worse performance from machines on a particular provider, then you can safely conclude that there is a temporary problem with only that provider, not your application. Load testing solutions that are locked into a single provider limit the test engineer's ability to conduct realistic, reliable, large-scale tests.

Secure Your Data
In internal pre-production testing, the data used is often fake - not actual customer or user information. Further, you can be reasonably assured that any real data used is safe because testing is being conducted within the firewall. This is not the case when testing from the cloud on production data. You'll need to take steps to ensure that any accounts, scenarios, detailed error messages, and other sensitive data involved in your tests are secured.

Encrypt the communication between your controller and load generators. This helps secure data sent to the load generators during the test (including account information) as well as the data that is retrieved (including error messages). If possible, use SSL to secure the communication between the browser and the tested server. Last, ensure your load generators are secured with their own firewalls to protect them from outside threats.

Tune Load Generators
To ensure that your load generator machines in the cloud are capable of generating large loads, you must properly tune the system to support the creation of a high number of sockets and threads per process. In addition, allocate an appropriate heap size for Java-based load generators. The default settings for a typical machine allow all programs to share its resources fairly. In the case of load generators, the machine is dedicated to a single task, so you can improve performance by allocating a significantly larger share of the available resources to the load generation tasks.

Monitor Your Servers
Once you've identified a performance bottleneck, you need information to track down its root cause. This information should be gathered during the test by monitoring each component of the infrastructure including application servers and database servers. Specifically, you want to monitor both the system - including the operating system, disks, and network - and the server software - including connection pools, threads, cache hits, and indexes.

Linking all the information gathered during the tests with the tests themselves is much easier when the monitoring is integrated with your load testing tool. This enables you to correlate the response times and errors generated by load testing with the monitored data to track down the cause of problems quickly. A cloud testing solution that has no ability to monitor activity inside the firewall cannot integrate and correlate the tests it initiates from the outside with what is happening on the inside. With such a setup, test engineers will not have all the information they need to quickly identify the sources of performance problems.

Summing It Up
Even with all its potential benefits, cloud testing cannot meet all the challenges facing performance test engineers. In practice, cloud testing is most effective when combined with internal load testing in a two-stage process that makes use of multiple cloud providers and both internal and external infrastructure. You can be optimistic about using the cloud, but don't get caught up in all the hype. The key is to think of cloud testing as the delivery mechanism that is just one (albeit an important one) part of an overall load and performance test strategy.

More Stories By Steve Weisfeldt

Steve Weisfeldt is a Senior Performance Engineer at Neotys, a provider of load testing software for Web applications. Previously, he has worked as the President of Engine 1 Consulting, a services firm specializing in all facets of test automation. Prior to his involvement at Engine 1 Consulting, he was a Senior Systems Engineer at Aternity. Prior to that, Steve spent seven years at automated testing vendor Segue Software (acquired by Borland). While spending most of his time at Segue delivering professional services and training, he was also involved in pre-sales and product marketing efforts.

Being in the load and performance testing space since 1999, Steve has been involved in load and performance testing projects of all sizes, in industries that span the retail, financial services, insurance and manufacturing sectors. His expertise lies in enabling organizations to optimize their ability to develop, test and launch high-quality applications efficiently, on-time and on-budget. Steve graduated from the University of Massachusetts-Lowell with a BS in Electrical Engineering and an MS in Computer Engineering.

Comments (0)

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.


@CloudExpo Stories
You use an agile process; your goal is to make your organization more agile. But what about your data infrastructure? The truth is, today's databases are anything but agile - they are effectively static repositories that are cumbersome to work with, difficult to change, and cannot keep pace with application demands. Performance suffers as a result, and it takes far longer than it should to deliver new features and capabilities needed to make your organization competitive. As your application an...
Want to enable self-service provisioning of application environments in minutes that mirror production? Can you automatically provide rich data with code-level detail back to the developers when issues occur in production? In his session at DevOps Summit, David Tesar, Microsoft Technical Evangelist on Microsoft Azure and DevOps, will discuss how to accomplish this and more utilizing technologies such as Microsoft Azure, Visual Studio online, and Application Insights in this demo-heavy session.
As cloud gives an opportunity to businesses to buy services externally – how is cloud impacting your customers? In his General Session at 15th Cloud Expo, Fabio Gori, Director of Worldwide Cloud Marketing at Cisco, provided answers to big questions: Do you see hybrid cloud as where the world is going? What benefits does it bring? And how does Cisco connect all of these clouds? He also discussed Intercloud and Cisco’s investment on it.
The Internet of Things is not new. Historically, smart businesses have used its basic concept of leveraging data to drive better decision making and have capitalized on those insights to realize additional revenue opportunities. So, what has changed to make the Internet of Things one of the hottest topics in tech? In his session at @ThingsExpo, Chris Gray, Director, Embedded and Internet of Things, discussed the underlying factors that are driving the economics of intelligent systems. Discover ...
Health care systems across the globe are under enormous strain, as facilities reach capacity and costs continue to rise. M2M and the Internet of Things have the potential to transform the industry through connected health solutions that can make care more efficient while reducing costs. In fact, Vodafone's annual M2M Barometer Report forecasts M2M applications rising to 57 percent in health care and life sciences by 2016. Lively is one of Vodafone's health care partners, whose solutions enable o...
In her General Session at 15th Cloud Expo, Anne Plese, Senior Consultant, Cloud Product Marketing, at Verizon Enterprise, focused on finding the right mix of renting vs. buying Oracle capacity to scale to meet business demands, and offer validated Oracle database TCO models for Oracle development and testing environments. Anne Plese is a marketing and technology enthusiast/realist with over 19+ years in high tech. At Verizon Enterprise, she focuses on driving growth for the Verizon Cloud platfo...
Andi Mann has been serving as Conference Chair of the DevOps Summit since its inception. He is one of the world's recognized leaders in DevOps, and continues to be one of its most articulate advocates. Here are some recent thoughts of his in an interview we conducted in the run-up to the DevOps Summit to be held June 9-11 at the Javits Center in New York City. When did you first start thinking about DevOps and its potential impact on enterprise IT? Andi: I first started thinking about DevOps b...
The most often asked question post-DevOps introduction is: “How do I get started?” There’s plenty of information on why DevOps is valid and important, but many managers still struggle with simple basics for how to initiate a DevOps program in their business. They struggle with issues related to current organizational inertia, the lack of experience on Continuous Integration/Delivery, understanding where DevOps will affect revenue and budget, etc. In their session at DevOps Summit, JP Morgentha...
In a recent research, analyst firm IDC found that the average cost of a critical application failure is $500,000 to $1 million per hour and the average total cost of unplanned application downtime is $1.25 billion to $2.5 billion per year for Fortune 1000 companies. In addition to the findings on the cost of the downtime, the research also highlighted best practices for development, testing, application support, infrastructure, and operations teams.
How does one bridge the gap between traditional enterprise storage infrastructures and the private, hybrid, and public cloud? In his session at 15th Cloud Expo, Dan Pollack, Chief Architect of Storage Operations at AOL Inc., examed the workload differences and required changes to reuse existing knowledge and components when building and using a cloud infrastructure. He also looked into the operational considerations, tool requirements, and behavioral changes required for private cloud storage s...
Software is eating the world. Companies that were not previously in the technology space now find themselves competing with Google and Amazon on speed of innovation. As the innovation cycle accelerates, companies must embrace rapid and constant change to both applications and their infrastructure, and find a way to deliver speed and agility of development without sacrificing reliability or efficiency of operations. In her Day 2 Keynote DevOps Summit, Victoria Livschitz, CEO of Qubell, discussed...
The speed of product development has increased massively in the past 10 years. At the same time our formal secure development and SDL methodologies have fallen behind. This forces product developers to choose between rapid release times and security. In his session at DevOps Summit, Michael Murray, Director of Cyber Security Consulting and Assessment at GE Healthcare, examined the problems and presented some solutions for moving security into the DevOps lifecycle to ensure that we get fast AND ...
SYS-CON Events announced today that MetraTech, now part of Ericsson, has been named “Silver Sponsor” of SYS-CON's 16th International Cloud Expo®, which will take place on June 9–11, 2015, at the Javits Center in New York, NY. Ericsson is the driving force behind the Networked Society- a world leader in communications infrastructure, software and services. Some 40% of the world’s mobile traffic runs through networks Ericsson has supplied, serving more than 2.5 billion subscribers.
The OpenStack cloud operating system includes Trove, a database abstraction layer. Rather than applications connecting directly to a specific type of database, they connect to Trove, which in turn connects to one or more specific databases. One target database is Postgres Plus Cloud Database, which includes its own RESTful API. Trove was originally developed around MySQL, whose interfaces are significantly less complicated than those of the Postgres cloud database. In his session at 16th Cloud...
Explosive growth in connected devices. Enormous amounts of data for collection and analysis. Critical use of data for split-second decision making and actionable information. All three are factors in making the Internet of Things a reality. Yet, any one factor would have an IT organization pondering its infrastructure strategy. How should your organization enhance its IT framework to enable an Internet of Things implementation? In this session, James Kirkland, Red Hat's Chief Architect for the ...
In their general session at 16th Cloud Expo, Michael Piccininni, Global Account Manager – Cloud SP at EMC Corporation, and Mike Dietze, Regional Director at Windstream Hosted Solutions, will review next generation cloud services, including the Windstream-EMC Tier Storage solutions, and discuss how to increase efficiencies, improve service delivery and enhance corporate cloud solution development. Speaker Bios Michael Piccininni is Global Account Manager – Cloud SP at EMC Corporation. He has b...
While there are hundreds of public and private cloud hosting providers to choose from, not all clouds are created equal. If you’re seeking to host enterprise-level mission-critical applications, where Cloud Security is a primary concern, WHOA.com is setting new standards for cloud hosting, and has established itself as a major contender in the marketplace. We are constantly seeking ways to innovate and leverage state-of-the-art technologies. In his session at 16th Cloud Expo, Mike Rivera, Seni...
Hardware will never be more valuable than on the day it hits your loading dock. Each day new servers are not deployed to production the business is losing money. While Moore's Law is typically cited to explain the exponential density growth of chips, a critical consequence of this is rapid depreciation of servers. The hardware for clustered systems (e.g., Hadoop, OpenStack) tends to be significant capital expenses. In his session at Big Data Expo, Mason Katz, CTO and co-founder of StackIQ, disc...
There is no question that the cloud is where businesses want to host data. Until recently hypervisor virtualization was the most widely used method in cloud computing. Recently virtual containers have been gaining in popularity, and for good reason. In the debate between virtual machines and containers, the latter have been seen as the new kid on the block – and like other emerging technology have had some initial shortcomings. However, the container space has evolved drastically since coming on...
The 4th International Internet of @ThingsExpo, co-located with the 17th International Cloud Expo - to be held November 3-5, 2015, at the Santa Clara Convention Center in Santa Clara, CA - announces that its Call for Papers is open. The Internet of Things (IoT) is the biggest idea since the creation of the Worldwide Web more than 20 years ago.