Welcome!

Microservices Expo Authors: Automic Blog, Elizabeth White, Harry Trott, Pat Romanski, Liz McMillan

Related Topics: Java IoT, Microservices Expo, Machine Learning , Agile Computing

Java IoT: Article

Don’t Let Load Balancers Ruin Your Holiday Business

Proactive performance management

An eCommerce site that crashes seven times during the Christmas season being down for up to five hours each time it crashes is a site that loses a lot of money and reputation. It happened to one of our customers who told this story at our annual performance conference earlier this month. Among the several reasons that led to these crashes I want to share more details on one of them that I see more often with other websites as well. Load balancers on a round-robin instead of least-busy can easily lead to app server crashes caused by heap memory exhaustion. Let's dig into some details on how to identify these problems and how to avoid it.

The Symptom: Crashing Tomcat Instances
The website is deployed on six Tomcats with three front-end Apache Web Servers. During peak load hours individual Tomcat instances started showing growing response times and a growing number of requests in the Tomcat processing queue. After a while these instances crashed due to out-of-memory exceptions and with that also brought down the rest of the site as load couldn't be handled any more with the remaining servers. Figure 1 shows the actual flow of transactions through the system highlighting unevenly distributed response time in the application servers and functional errors being reported on all tiers (red-colored server icon):

Even with equally distributed load (Round Robin Load Balancer Setting) one of the Tomcats spiked in Response Time Contribution before crashing

Once the App Server started rejecting incoming connections we can observe the first ripple effect of errors. We can see a very high number of exceptions in the database layer, exceptions thrown between application tiers with the web app responding with HTTP 500s:

Within 30 minutes the application serves 43000 pages with an HTTP 500 Response correlating to Exceptions in the Database and Inter-Tier Communication

The Root Cause: Inefficient Database Statements and Connection Pool Usage
The exceptions caught in the Database Layer (JDBC) were already a very good hint of the root cause of this problem. A closer look at the Exceptions shows that connection pools are exhausted, which causes problems in the different components of the application:

Exhausted Connection Pool causes Exceptions that impact Data Access Layer as well as Widget Rendering

Looking at the performance breakdown by application layer reveals how much performance impact connection pooling has on the overall transaction response time:

Due to the connection pool problem a single request had to wait 3.8s on average to obtain a connection from the pool

Now - it was not only the size of the pool that was the problem - but - several very inefficient database statements that took long to execute for certain business transactions of the application. This caused the application server to hold on to the connection longer than normal. As the load balancer was configured with Round Robin the app server still got additional requests served. Eventually - just by the random nature of incoming requests - one app server received several of these requests that executed these inefficient database calls. Once the connection pool was exhausted the application started throwing exceptions that ultimately also led to a crash of the JVM. Once the first app server crashed, it didn't take too long to take the other app servers down as well.

The Solution: Optimizing App and Load Balancer
The problem was fixed by looking at the slowest database statements and optimizing them for performance by, e.g, adding indices on the database or making the SQL statements more efficient. They also optimized the pool size to accommodate the expected load during peak hours.

 

They started by optimizing SQL Statements that took long to execute and those that got executed several times within the same transaction

They also changed the load balancer setting from Round-Robin to Least-Busy, which was the preferred setting from the LB vendor but had simply forgotten to configure in the production environment.

The Result: Site Has Not Been Down Since
Since they made the changes to the application and the load balancer the site has never gone down since. Now - the next holiday season is coming up and they are ready for the next seasonal spikes. Even though they are really confident that everything will work without any problems, they learned their lesson and are approaching performance proactively through proper load testing.

Next Steps: Proactive Performance Management
The lesson learned was that these problems could have been found prior to the holiday shopping season by doing proper load testing. They did load testing before but never encountered this problem because of two reasons: 1) they didn't test using expected peak volumes for long enough sessions and 2) didn't use a tool that simulated real customer behavior variations (too few scripts and the scripts were too simple) that tested their highly interactive web site.

Their strategy for proactive performance management is that they:

  1. Perform load tests on the production system during low traffic hours (2 a.m.-6 a.m.), accepting the risk of minor sales losses in case of a crash, versus major sales losses during the holiday shopping season.
  2. Multiply the hourly load test volume by 2.5 since their actual peaks are 10 hours long.
  3. Use a load testing service that uses real browsers in different locations around the U.S.
  4. Use an APM solution that identifies problems within the application while running the load test.

If you want to read more on common performance problems that are not found prior to moving to production check out my recent series of blogs: Supersized Content, Deployment Mistakes or Excessive Logging

More Stories By Michael Kopp

Michael Kopp has over 12 years of experience as an architect and developer in the Enterprise Java space. Before coming to CompuwareAPM dynaTrace he was the Chief Architect at GoldenSource, a major player in the EDM space. In 2009 he joined dynaTrace as a technology strategist in the center of excellence. He specializes application performance management in large scale production environments with special focus on virtualized and cloud environments. His current focus is how to effectively leverage BigData Solutions and how these technologies impact and change the application landscape.

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.


@MicroservicesExpo Stories
Docker is on a roll. In the last few years, this container management service has become immensely popular in development, especially given the great fit with agile-based projects and continuous delivery. In this article, I want to take a brief look at how you can use Docker to accelerate and streamline the software development lifecycle (SDLC) process.
While some vendors scramble to create and sell you a fancy solution for monitoring your spanking new Amazon Lambdas, hear how you can do it on the cheap using just built-in Java APIs yourself. By exploiting a little-known fact that Lambdas aren’t exactly single-threaded, you can effectively identify hot spots in your serverless code. In his session at @DevOpsSummit at 21st Cloud Expo, Dave Martin, Product owner at CA Technologies, will give a live demonstration and code walkthrough, showing how ...
Cloud adoption is often driven by a desire to increase efficiency, boost agility and save money. All too often, however, the reality involves unpredictable cost spikes and lack of oversight due to resource limitations. In his session at 20th Cloud Expo, Joe Kinsella, CTO and Founder of CloudHealth Technologies, tackled the question: “How do you build a fully optimized cloud?” He will examine: Why TCO is critical to achieving cloud success – and why attendees should be thinking holistically ab...
There are several reasons why businesses migrate their operations to the cloud. Scalability and price are among the most important factors determining this transition. Unlike legacy systems, cloud based businesses can scale on demand. The database and applications in the cloud are not rendered simply from one server located in your headquarters, but is instead distributed across several servers across the world. Such CDNs also bring about greater control in times of uncertainty. A database hack ...
Did you know that you can develop for mainframes in Java? Or that the testing and deployment can be automated across mobile to mainframe? In his session and demo at @DevOpsSummit at 21st Cloud Expo, Dana Boudreau, a Senior Director at CA Technologies, will discuss how increasingly teams are developing with agile methodologies, using modern development environments, and automating testing and deployments, mobile to mainframe.
As DevOps methodologies expand their reach across the enterprise, organizations face the daunting challenge of adapting related cloud strategies to ensure optimal alignment, from managing complexity to ensuring proper governance. How can culture, automation, legacy apps and even budget be reexamined to enable this ongoing shift within the modern software factory?
With Cloud Foundry you can easily deploy and use apps utilizing websocket technology, but not everybody realizes that scaling them out is not that trivial. In his session at 21st Cloud Expo, Roman Swoszowski, CTO and VP, Cloud Foundry Services, at Grape Up, will show you an example of how to deal with this issue. He will demonstrate a cloud-native Spring Boot app running in Cloud Foundry and communicating with clients over websocket protocol that can be easily scaled horizontally and coordinate...
@DevOpsSummit at Cloud Expo taking place Oct 31 - Nov 2, 2017, at the Santa Clara Convention Center, Santa Clara, CA, is co-located with the 21st International Cloud Expo and will feature technical sessions from a rock star conference faculty and the leading industry players in the world. The widespread success of cloud computing is driving the DevOps revolution in enterprise IT. Now as never before, development teams must communicate and collaborate in a dynamic, 24/7/365 environment. There is ...
If you cannot explicitly articulate how investing in a new technology, changing the approach or re-engineering the business process will help you achieve your customer-centric vision of the future in direct and measurable ways, you probably shouldn’t be doing it. At Intellyx, we spend a lot of time talking to technology vendors. In our conversations, we explore emerging new technologies that are either disrupting the way enterprise organizations work or that help enable those organizations to co...
DevOps at Cloud Expo, taking place October 31 - November 2, 2017, at the Santa Clara Convention Center in Santa Clara, CA, is co-located with 21st Cloud Expo and will feature technical sessions from a rock star conference faculty and the leading industry players in the world. The widespread success of cloud computing is driving the DevOps revolution in enterprise IT. Now as never before, development teams must communicate and collaborate in a dynamic, 24/7/365 environment. There is no time to w...
In his session at 20th Cloud Expo, Scott Davis, CTO of Embotics, discussed how automation can provide the dynamic management required to cost-effectively deliver microservices and container solutions at scale. He also discussed how flexible automation is the key to effectively bridging and seamlessly coordinating both IT and developer needs for component orchestration across disparate clouds – an increasingly important requirement at today’s multi-cloud enterprise.
We define Hybrid IT as a management approach in which organizations create a workload-centric and value-driven integrated technology stack that may include legacy infrastructure, web-scale architectures, private cloud implementations along with public cloud platforms ranging from Infrastructure-as-a-Service to Software-as-a-Service.
IT organizations are moving to the cloud in hopes to approve efficiency, increase agility and save money. Migrating workloads might seem like a simple task, but what many businesses don’t realize is that application migration criteria differs across organizations, making it difficult for architects to arrive at an accurate TCO number. In his session at 21st Cloud Expo, Joe Kinsella, CTO of CloudHealth Technologies, will offer a systematic approach to understanding the TCO of a cloud application...
API Security has finally entered our security zeitgeist. OWASP Top 10 2017 - RC1 recognized API Security as a first class citizen by adding it as number 10, or A-10 on its list of web application vulnerabilities. We believe this is just the start. The attack surface area offered by API is orders or magnitude larger than any other attack surface area. Consider the fact the APIs expose cloud services, internal databases, application and even legacy mainframes over the internet. What could go wrong...
The goal of Continuous Testing is to shift testing left to find defects earlier and release software faster. This can be achieved by integrating a set of open source functional and performance testing tools in the early stages of your software delivery lifecycle. There is one process that binds all application delivery stages together into one well-orchestrated machine: Continuous Testing. Continuous Testing is the conveyer belt between the Software Factory and production stages. Artifacts are m...
In IT, we sometimes coin terms for things before we know exactly what they are and how they’ll be used. The resulting terms may capture a common set of aspirations and goals – as “cloud” did broadly for on-demand, self-service, and flexible computing. But such a term can also lump together diverse and even competing practices, technologies, and priorities to the point where important distinctions are glossed over and lost.
Most companies are adopting or evaluating container technology - Docker in particular - to speed up application deployment, drive down cost, ease management and make application delivery more flexible overall. As with most new architectures, this dream takes a lot of work to become a reality. Even when you do get your application componentized enough and packaged properly, there are still challenges for DevOps teams to making the shift to continuous delivery and achieving that reduction in cost ...
Enterprise architects are increasingly adopting multi-cloud strategies as they seek to utilize existing data center assets, leverage the advantages of cloud computing and avoid cloud vendor lock-in. This requires a globally aware traffic management strategy that can monitor infrastructure health across data centers and end-user experience globally, while responding to control changes and system specification at the speed of today’s DevOps teams. In his session at 20th Cloud Expo, Josh Gray, Chie...
"At the keynote this morning we spoke about the value proposition of Nutanix, of having a DevOps culture and a mindset, and the business outcomes of achieving agility and scale, which everybody here is trying to accomplish," noted Mark Lavi, DevOps Solution Architect at Nutanix, in this SYS-CON.tv interview at @DevOpsSummit at 20th Cloud Expo, held June 6-8, 2017, at the Javits Center in New York City, NY.
We have already established the importance of APIs in today’s digital world (read about it here). With APIs playing such an important role in keeping us connected, it’s necessary to maintain the API’s performance as well as availability. There are multiple aspects to consider when monitoring APIs, from integration to performance issues, therefore a general monitoring strategy that only accounts for up-time is not ideal.