Click here to close now.


@CloudExpo Authors: Elizabeth White, Liz McMillan, Pat Romanski, Gregor Petri, Ed Featherston

Related Topics: Java IoT, Microservices Expo, Adobe Flex, IoT User Interface, Apache

Java IoT: Article

Why Averages Are Inadequate, and Percentiles Are Great

Averages are ineffective because they are too simplistic and one-dimensional

Anyone who ever monitored or analyzed an application uses or has used averages. They are simple to understand and calculate. We tend to ignore just how wrong the picture is that averages paint of the world. To emphasis the point let me give you a real-world example outside of the performance space that I read recently in a newspaper.

The article was explaining that the average salary in a certain region in Europe was 1900 Euro's (to be clear this would be quite good in that region!). However when looking closer they found out that the majority, namely 9 out of 10 people, only earned around 1000 Euros and one would earn 10.000 (I over simplified this of course, but you get the idea). If you do the math you will see that the average of this is indeed 1900, but we can all agree that this does not represent the "average" salary as we would use the word in day to day live. So now let's apply this thinking to application performance.

The Average Response Time
The average response time is by far the most commonly used metric in application performance management. We assume that this represents a "normal" transaction, however this would only be true if the response time is always the same (all transaction run at equal speed) or the response time distribution is roughly bell curved.

A Bell curve represents the "normal" distribution of response times in which the average and the median are the same. It rarely ever occurs in real applications

In a Bell Curve the average (mean) and median are the same. In other words observed performance would represent the majority (half or more than half) of the transactions.

In reality most applications have few very heavy outliers; a statistician would say that the curve has a long tail. A long tail does not imply many slow transactions, but few that are magnitudes slower than the norm.

This is a typical Response Time Distribution with few but heavy outliers - it has a long tail. The average here is dragged to the right by the long tail.

We recognize that the average no longer represents the bulk of the transactions but can be a lot higher than the median.

You can now argue that this is not a problem as long as the average doesn't look better than the median. I would disagree, but let's look at another real-world scenario experienced by many of our customers:

This is another typical Response Time Distribution. Here we have quite a few very fast transactions that drag the average to the left of the actual median

In this case a considerable percentage of transactions are very, very fast (10-20 percent), while the bulk of transactions are several times slower. The median would still tell us the true story, but the average all of a sudden looks a lot faster than most of our transactions actually are. This is very typical in search engines or when caches are involved - some transactions are very fast, but the bulk are normal. Another reason for this scenario are failed transactions, more specifically transactions that failed fast. Many real-world applications have a failure rate of 1-10 percent (due to user errors or validation errors). These failed transactions are often magnitudes faster than the real ones and consequently distorted an average.

Of course performance analysts are not stupid and regularly try to compensate with higher frequency charts (compensating by looking at smaller aggregates visually) and by taking in minimum and maximum observed response times. However we can often only do this if we know the application very well, those unfamiliar with the application might easily misinterpret the charts. Because of the depth and type of knowledge required for this, it's difficult to communicate your analysis to other people - think how many arguments between IT teams have been caused by this. And that's before we even begin to think about communicating with business stakeholders!

A better metric by far are percentiles, because they allow us to understand the distribution. But before we look at percentiles, let's take a look a key feature in every production monitoring solution: Automatic Baselining and Alerting.

Automatic Baselining and Alerting
In real-world environments, performance gets attention when it is poor and has a negative impact on the business and users. But how can we identify performance issues quickly to prevent negative effects? We cannot alert on every slow transaction, since there are always some. In addition, most operations teams have to maintain a large number of applications and are not familiar with all of them, so manually setting thresholds can be inaccurate, quite painful and time-consuming.

The industry has come up with a solution called Automatic Baselining. Baselining calculates out the "normal" performance and only alerts us when an application slows down or produces more errors than usual. Most approaches rely on averages and standard deviations.

Without going into statistical details, this approach again assumes that the response times are distributed over a bell curve:

The Standard Deviation represents 33% of all transactions with the mean as the middle. 2xStandard Deviation represents 66% and thus the majority, everything outside could be considered an outlier. However most real world scenarios are not bell curved...

Typically, transactions that are outside two times standard deviation are treated as slow and captured for analysis. An alert is raised if the average moves significantly. In a bell curve this would account for the slowest 16.5 percent (and you can of course adjust that); however; if the response time distribution does not represent a bell curve, it becomes inaccurate. We either end up with a lot of false positives (transactions that are a lot slower than the average but when looking at the curve lie within the norm) or we miss a lot of problems (false negatives). In addition if the curve is not a bell curve, then the average can differ a lot from the median; applying a standard deviation to such an average can lead to quite a different result than you would expect. To work around this problem these algorithms have many tunable variables and a lot of "hacks" for specific use cases.

Why I Love Percentiles
A percentile tells me which part of the curve I am looking at and how many transactions are represented by that metric. To visualize this look at the following chart:

This chart shows the 50th and 90th percentile along with the average of the same transaction. It shows that the average is influenced far mor heavily by the 90th, thus by outliers and not by the bulk of the transactions

The green line represents the average. As you can see it is very volatile. The other two lines represent the 50th and 90th percentile. As we can see the 50th percentile (or median) is rather stable but has a couple of jumps. These jumps represent real performance degradation for the majority (50%) of the transactions. The 90th percentile (this is the start of the "tail") is a lot more volatile, which means that the outliers slowness depends on data or user behavior. What's important here is that the average is heavily influenced (dragged) by the 90th percentile, the tail, rather than the bulk of the transactions.

If the 50th percentile (median) of a response time is 500ms that means that 50% of my transactions are either as fast or faster than 500ms. If the 90th percentile of the same transaction is at 1000ms it means that 90% are as fast or faster and only 10% are slower. The average in this case could either be lower than 500ms (on a heavy front curve), a lot higher (long tail) or somewhere in between. A percentile gives me a much better sense of my real world performance, because it shows me a slice of my response time curve.

For exactly that reason percentiles are perfect for automatic baselining. If the 50th percentile moves from 500ms to 600ms I know that 50% of my transactions suffered a 20% performance degradation. You need to react to that.

In many cases we see that the 75th or 90th percentile does not change at all in such a scenario. This means the slow transactions didn't get any slower, only the normal ones did. Depending on how long your tail is the average might not have moved at all in such a scenario.!

In other cases we see the 98th percentile degrading from 1s to 1.5 seconds while the 95th is stable at 900ms. This means that your application as a whole is stable, but a few outliers got worse, nothing to worry about immediately. Percentile-based alerts do not suffer from false positives, are a lot less volatile and don't miss any important performance degradations! Consequently a baselining approach that uses percentiles does not require a lot of tuning variables to work effectively.

The screenshot below shows the Median (50th Percentile) for a particular transaction jumping from about 50ms to about 500ms and triggering an alert as it is significantly above the calculated baseline (green line). The chart labeled "Slow Response Time" on the other hand shows the 90th percentile for the same transaction. These "outliers" also show an increase in response time but not significant enough to trigger an alert.

Here we see an automatic baselining dashboard with a violation at the 50th percentile. The violation is quite clear, at the same time the 90th percentile (right upper chart) does not violate. Because the outliers are so much slower than the bulk of the transaction an average would have been influenced by them and would not have have reacted quite as dramatically as the 50th percentile. We might have missed this clear violation!

How Can We Use Percentiles for Tuning?
Percentiles are also great for tuning, and giving your optimizations a particular goal. Let's say that something within my application is too slow in general and I need to make it faster. In this case I want to focus on bringing down the 90th percentile. This would ensure sure that the overall response time of the application goes down. In other cases I have unacceptably long outliers I want to focus on bringing down response time for transactions beyond the 98th or 99th percentile (only outliers). We see a lot of applications that have perfectly acceptable performance for the 90th percentile, with the 98th percentile being magnitudes worse.

In throughput oriented applications on the other hand I would want to make the majority of my transactions very fast, while accepting that an optimization makes a few outliers slower. I might therefore make sure that the 75th percentile goes down while trying to keep the 90th percentile stable or not getting a lot worse.

I could not make the same kind of observations with averages, minimum and maximum, but with percentiles they are very easy indeed.

Averages are ineffective because they are too simplistic and one-dimensional. Percentiles are a really great and easy way of understanding the real performance characteristics of your application. They also provide a great basis for automatic baselining, behavioral learning and optimizing your application with a proper focus. In short, percentiles are great!

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 (1) View Comments

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.

Most Recent Comments
rtalexander 11/21/12 12:58:00 AM EST

Hey, could you post a reference or two that covers the theory and/or practicalities of the approach you describe?


@CloudExpo Stories
The buzz continues for cloud, data analytics and the Internet of Things (IoT) and their collective impact across all industries. But a new conversation is emerging - how do companies use industry disruption and technology enablers to lead in markets undergoing change, uncertainty and ambiguity? Organizations of all sizes need to evolve and transform, often under massive pressure, as industry lines blur and merge and traditional business models are assaulted and turned upside down. In this new da...
The Internet of Things (IoT) is growing rapidly by extending current technologies, products and networks. By 2020, Cisco estimates there will be 50 billion connected devices. Gartner has forecast revenues of over $300 billion, just to IoT suppliers. Now is the time to figure out how you’ll make money – not just create innovative products. With hundreds of new products and companies jumping into the IoT fray every month, there’s no shortage of innovation. Despite this, McKinsey/VisionMobile data...
Containers are changing the security landscape for software development and deployment. As with any security solutions, security approaches that work for developers, operations personnel and security professionals is a requirement. In his session at @DevOpsSummit, Kevin Gilpin, CTO and Co-Founder of Conjur, will discuss various security considerations for container-based infrastructure and related DevOps workflows.
As-a-service models offer huge opportunities, but also complicate security. It may seem that the easiest way to migrate to a new architectural model is to let others, experts in their field, do the work. This has given rise to many as-a-service models throughout the industry and across the entire technology stack, from software to infrastructure. While this has unlocked huge opportunities to accelerate the deployment of new capabilities or increase economic efficiencies within an organization, i...
There are so many tools and techniques for data analytics that even for a data scientist the choices, possible systems, and even the types of data can be daunting. In his session at @ThingsExpo, Chris Harrold, Global CTO for Big Data Solutions for EMC Corporation, will show how to perform a simple, but meaningful analysis of social sentiment data using freely available tools that take only minutes to download and install. Participants will get the download information, scripts, and complete en...
Containers are revolutionizing the way we deploy and maintain our infrastructures, but monitoring and troubleshooting in a containerized environment can still be painful and impractical. Understanding even basic resource usage is difficult - let alone tracking network connections or malicious activity. In his session at DevOps Summit, Gianluca Borello, Sr. Software Engineer at Sysdig, will cover the current state of the art for container monitoring and visibility, including pros / cons and li...
IT data is typically silo'd by the various tools in place. Unifying all the log, metric and event data in one analytics platform stops finger pointing and provides the end-to-end correlation. Logs, metrics and custom event data can be joined to tell the holistic story of your software and operations. For example, users can correlate code deploys to system performance to application error codes.
Between the compelling mockups and specs produced by analysts, and resulting applications built by developers, there exists a gulf where projects fail, costs spiral, and applications disappoint. Methodologies like Agile attempt to address this with intensified communication, with partial success but many limitations. In his session at DevOps Summit, Charles Kendrick, CTO and Chief Architect at Isomorphic Software, will present a revolutionary model enabled by new technologies. Learn how busine...
There are many considerations when moving applications from on-premise to cloud. It is critical to understand the benefits and also challenges of this migration. A successful migration will result in lower Total Cost of Ownership, yet offer the same or higher level of robustness. Migration to cloud shifts computing resources from your data center, which can yield significant advantages provided that the cloud vendor an offer enterprise-grade quality for your application.
Internet of Things (IoT) will be a hybrid ecosystem of diverse devices and sensors collaborating with operational and enterprise systems to create the next big application. In their session at @ThingsExpo, Bramh Gupta, founder and CEO of, and Fred Yatzeck, principal architect leading product development at, discussed how choosing the right middleware and integration strategy from the get-go will enable IoT solution developers to adapt and grow with the industry, while at th...
Manufacturing has widely adopted standardized and automated processes to create designs, build them, and maintain them through their life cycle. However, many modern manufacturing systems go beyond mechanized workflows to introduce empowered workers, flexible collaboration, and rapid iteration. Such behaviors also characterize open source software development and are at the heart of DevOps culture, processes, and tooling.
The web app is agile. The REST API is agile. The testing and planning are agile. But alas, data infrastructures certainly are not. Once an application matures, changing the shape or indexing scheme of data often forces at best a top down planning exercise and at worst includes schema changes that force downtime. The time has come for a new approach that fundamentally advances the agility of distributed data infrastructures. Come learn about a new solution to the problems faced by software organ...
The last decade was about virtual machines, but the next one is about containers. Containers enable a service to run on any host at any time. Traditional tools are starting to show cracks because they were not designed for this level of application portability. Now is the time to look at new ways to deploy and manage applications at scale. In his session at @DevOpsSummit, Brian “Redbeard” Harrington, a principal architect at CoreOS, will examine how CoreOS helps teams run in production. Attende...
“All our customers are looking at the cloud ecosystem as an important part of their overall product strategy. Some see it evolve as a multi-cloud / hybrid cloud strategy, while others are embracing all forms of cloud offerings like PaaS, IaaS and SaaS in their solutions,” noted Suhas Joshi, Vice President – Technology, at Harbinger Group, in this exclusive Q&A with Cloud Expo Conference Chair Roger Strukhoff.
Can call centers hang up the phones for good? Intuitive Solutions did. WebRTC enabled this contact center provider to eliminate antiquated telephony and desktop phone infrastructure with a pure web-based solution, allowing them to expand beyond brick-and-mortar confines to a home-based agent model. It also ensured scalability and better service for customers, including MUY! Companies, one of the country's largest franchise restaurant companies with 232 Pizza Hut locations. This is one example of...
Cloud computing delivers on-demand resources that provide businesses with flexibility and cost-savings. The challenge in moving workloads to the cloud has been the cost and complexity of ensuring the initial and ongoing security and regulatory (PCI, HIPAA, FFIEC) compliance across private and public clouds. Manual security compliance is slow, prone to human error, and represents over 50% of the cost of managing cloud applications. Determining how to automate cloud security compliance is critical...
SYS-CON Events announced today that VividCortex, the monitoring solution for the modern data system, will exhibit at the 17th International Cloud Expo®, which will take place on November 3–5, 2015, at the Santa Clara Convention Center in Santa Clara, CA. The database is the heart of most applications, but it’s also the part that’s hardest to scale, monitor, and optimize even as it’s growing 50% year over year. VividCortex is the first unified suite of database monitoring tools specifically desi...
Saviynt Inc. has announced the availability of the next release of Saviynt for AWS. The comprehensive security and compliance solution provides a Command-and-Control center to gain visibility into risks in AWS, enforce real-time protection of critical workloads as well as data and automate access life-cycle governance. The solution enables AWS customers to meet their compliance mandates such as ITAR, SOX, PCI, etc. by including an extensive risk and controls library to detect known threats and b...
You have your devices and your data, but what about the rest of your Internet of Things story? Two popular classes of technologies that nicely handle the Big Data analytics for Internet of Things are Apache Hadoop and NoSQL. Hadoop is designed for parallelizing analytical work across many servers and is ideal for the massive data volumes you create with IoT devices. NoSQL databases such as Apache HBase are ideal for storing and retrieving IoT data as “time series data.”
Clearly the way forward is to move to cloud be it bare metal, VMs or containers. One aspect of the current public clouds that is slowing this cloud migration is cloud lock-in. Every cloud vendor is trying to make it very difficult to move out once a customer has chosen their cloud. In his session at 17th Cloud Expo, Naveen Nimmu, CEO of Clouber, Inc., will advocate that making the inter-cloud migration as simple as changing airlines would help the entire industry to quickly adopt the cloud wit...