Welcome!

Microsoft Cloud Authors: Pat Romanski, Elizabeth White, Liz McMillan, Mihai Corbuleac, David Bermingham

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We recognize that the average no longer represents the bulk of the transactions but can be a lot higher than the median.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

Conclusion
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?

Thanks!

@ThingsExpo Stories
It is one thing to build single industrial IoT applications, but what will it take to build the Smart Cities and truly society changing applications of the future? The technology won’t be the problem, it will be the number of parties that need to work together and be aligned in their motivation to succeed. In his Day 2 Keynote at @ThingsExpo, Henrik Kenani Dahlgren, Portfolio Marketing Manager at Ericsson, discussed how to plan to cooperate, partner, and form lasting all-star teams to change t...
Connected devices and the industrial internet are growing exponentially every year with Cisco expecting 50 billion devices to be in operation by 2020. In this period of growth, location-based insights are becoming invaluable to many businesses as they adopt new connected technologies. Knowing when and where these devices connect from is critical for a number of scenarios in supply chain management, disaster management, emergency response, M2M, location marketing and more. In his session at @Th...
Internet of @ThingsExpo, taking place November 1-3, 2016, at the Santa Clara Convention Center in Santa Clara, CA, is co-located with the 19th International Cloud Expo and will feature technical sessions from a rock star conference faculty and the leading industry players in the world and ThingsExpo Silicon Valley Call for Papers is now open.
Machine Learning helps make complex systems more efficient. By applying advanced Machine Learning techniques such as Cognitive Fingerprinting, wind project operators can utilize these tools to learn from collected data, detect regular patterns, and optimize their own operations. In his session at 18th Cloud Expo, Stuart Gillen, Director of Business Development at SparkCognition, discussed how research has demonstrated the value of Machine Learning in delivering next generation analytics to imp...
SYS-CON Events announced today that ReadyTalk, a leading provider of online conferencing and webinar services, has been named Vendor Presentation Sponsor at the 19th International Cloud Expo, which will take place on November 1–3, 2016, at the Santa Clara Convention Center in Santa Clara, CA. ReadyTalk delivers audio and web conferencing services that inspire collaboration and enable the Future of Work for today’s increasingly digital and mobile workforce. By combining intuitive, innovative tec...
Amazon has gradually rolled out parts of its IoT offerings, but these are just the tip of the iceberg. In addition to optimizing their backend AWS offerings, Amazon is laying the ground work to be a major force in IoT - especially in the connected home and office. In his session at @ThingsExpo, Chris Kocher, founder and managing director of Grey Heron, explained how Amazon is extending its reach to become a major force in IoT by building on its dominant cloud IoT platform, its Dash Button strat...
The cloud market growth today is largely in public clouds. While there is a lot of spend in IT departments in virtualization, these aren’t yet translating into a true “cloud” experience within the enterprise. What is stopping the growth of the “private cloud” market? In his general session at 18th Cloud Expo, Nara Rajagopalan, CEO of Accelerite, explored the challenges in deploying, managing, and getting adoption for a private cloud within an enterprise. What are the key differences between wh...
In his keynote at 18th Cloud Expo, Andrew Keys, Co-Founder of ConsenSys Enterprise, provided an overview of the evolution of the Internet and the Database and the future of their combination – the Blockchain. Andrew Keys is Co-Founder of ConsenSys Enterprise. He comes to ConsenSys Enterprise with capital markets, technology and entrepreneurial experience. Previously, he worked for UBS investment bank in equities analysis. Later, he was responsible for the creation and distribution of life sett...
19th Cloud Expo, taking place November 1-3, 2016, at the Santa Clara Convention Center in Santa Clara, CA, will feature technical sessions from a rock star conference faculty and the leading industry players in the world. Cloud computing is now being embraced by a majority of enterprises of all sizes. Yesterday's debate about public vs. private has transformed into the reality of hybrid cloud: a recent survey shows that 74% of enterprises have a hybrid cloud strategy. Meanwhile, 94% of enterpri...
SYS-CON Events announced today that Bsquare has been named “Silver Sponsor” of SYS-CON's @ThingsExpo, which will take place on November 1–3, 2016, at the Santa Clara Convention Center in Santa Clara, CA. For more than two decades, Bsquare has helped its customers extract business value from a broad array of physical assets by making them intelligent, connecting them, and using the data they generate to optimize business processes.
There are several IoTs: the Industrial Internet, Consumer Wearables, Wearables and Healthcare, Supply Chains, and the movement toward Smart Grids, Cities, Regions, and Nations. There are competing communications standards every step of the way, a bewildering array of sensors and devices, and an entire world of competing data analytics platforms. To some this appears to be chaos. In this power panel at @ThingsExpo, moderated by Conference Chair Roger Strukhoff, Bradley Holt, Developer Advocate a...
There is little doubt that Big Data solutions will have an increasing role in the Enterprise IT mainstream over time. Big Data at Cloud Expo - to be held November 1-3, 2016, at the Santa Clara Convention Center in Santa Clara, CA - has announced its Call for Papers is open. Cloud computing is being adopted in one form or another by 94% of enterprises today. Tens of billions of new devices are being connected to The Internet of Things. And Big Data is driving this bus. An exponential increase is...
Internet of @ThingsExpo, taking place November 1-3, 2016, at the Santa Clara Convention Center in Santa Clara, CA, is co-located with 19th Cloud Expo and will feature technical sessions from a rock star conference faculty and the leading industry players in the world. The Internet of Things (IoT) is the most profound change in personal and enterprise IT since the creation of the Worldwide Web more than 20 years ago. All major researchers estimate there will be tens of billions devices - comp...
A strange thing is happening along the way to the Internet of Things, namely far too many devices to work with and manage. It has become clear that we'll need much higher efficiency user experiences that can allow us to more easily and scalably work with the thousands of devices that will soon be in each of our lives. Enter the conversational interface revolution, combining bots we can literally talk with, gesture to, and even direct with our thoughts, with embedded artificial intelligence, wh...
Cloud computing is being adopted in one form or another by 94% of enterprises today. Tens of billions of new devices are being connected to The Internet of Things. And Big Data is driving this bus. An exponential increase is expected in the amount of information being processed, managed, analyzed, and acted upon by enterprise IT. This amazing is not part of some distant future - it is happening today. One report shows a 650% increase in enterprise data by 2020. Other estimates are even higher....
Cognitive Computing is becoming the foundation for a new generation of solutions that have the potential to transform business. Unlike traditional approaches to building solutions, a cognitive computing approach allows the data to help determine the way applications are designed. This contrasts with conventional software development that begins with defining logic based on the current way a business operates. In her session at 18th Cloud Expo, Judith S. Hurwitz, President and CEO of Hurwitz & ...
In his general session at 18th Cloud Expo, Lee Atchison, Principal Cloud Architect and Advocate at New Relic, discussed cloud as a ‘better data center’ and how it adds new capacity (faster) and improves application availability (redundancy). The cloud is a ‘Dynamic Tool for Dynamic Apps’ and resource allocation is an integral part of your application architecture, so use only the resources you need and allocate /de-allocate resources on the fly.
The 19th International Cloud Expo has announced that its Call for Papers is open. Cloud Expo, to be held November 1-3, 2016, at the Santa Clara Convention Center in Santa Clara, CA, brings together Cloud Computing, Big Data, Internet of Things, DevOps, Digital Transformation, Microservices and WebRTC to one location. With cloud computing driving a higher percentage of enterprise IT budgets every year, it becomes increasingly important to plant your flag in this fast-expanding business opportuni...
industrial company for a multi-year contract initially valued at over $4.0 million. In addition to DataV software, Bsquare will also provide comprehensive systems integration, support and maintenance services. DataV leverages advanced data analytics, predictive reasoning, data-driven diagnostics, and automated orchestration of remediation actions in order to improve asset uptime while reducing service and warranty costs.
Vidyo, Inc., has joined the Alliance for Open Media. The Alliance for Open Media is a non-profit organization working to define and develop media technologies that address the need for an open standard for video compression and delivery over the web. As a member of the Alliance, Vidyo will collaborate with industry leaders in pursuit of an open and royalty-free AOMedia Video codec, AV1. Vidyo’s contributions to the organization will bring to bear its long history of expertise in codec technolo...