Click here to close now.

Welcome!

Microservices Journal Authors: Liz McMillan, David Sprott, Lori MacVittie, Carmen Gonzalez, Elizabeth White

Related Topics: Java, Microservices Journal, Adobe Flex, AJAX & REA, Apache

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We recognize that the average no longer represents the bulk of the transactions but can be a lot higher than the median.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

@MicroservicesExpo Stories
There are 182 billion emails sent every day, generating a lot of data about how recipients and ISPs respond. Many marketers take a more-is-better approach to stats, preferring to have the ability to slice and dice their email lists based numerous arbitrary stats. However, fundamentally what really matters is whether or not sending an email to a particular recipient will generate value. Data Scientists can design high-level insights such as engagement prediction models and content clusters that a...
For those of us that have been practicing SOA for over a decade, it's surprising that there's so much interest in microservices. In fairness microservices don't look like the vendor play that was early SOA in the early noughties. But experienced SOA practitioners everywhere will be wondering if microservices is actually a good thing. You see microservices is basically an SOA pattern that inherits all the well-known SOA principles and adds characteristics that address the use of SOA for distribut...
Microservices are the result of decomposing applications. That may sound a lot like SOA, but SOA was based on an object-oriented (noun) premise; that is, services were built around an object - like a customer - with all the necessary operations (functions) that go along with it. SOA was also founded on a variety of standards (most of them coming out of OASIS) like SOAP, WSDL, XML and UDDI. Microservices have no standards (at least none deriving from a standards body or organization) and can be b...
17th Cloud Expo, taking place Nov 3-5, 2015, 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 enterprises a...
In today's digital world, change is the one constant. Disruptive innovations like cloud, mobility, social media, and the Internet of Things have reshaped the market and set new standards in customer expectations. To remain competitive, businesses must tap the potential of emerging technologies and markets through the rapid release of new products and services. However, the rigid and siloed structures of traditional IT platforms and processes are slowing them down – resulting in lengthy delivery ...
In today's application economy, enterprise organizations realize that it's their applications that are the heart and soul of their business. If their application users have a bad experience, their revenue and reputation are at stake. In his session at 15th Cloud Expo, Anand Akela, Senior Director of Product Marketing for Application Performance Management at CA Technologies, discussed how a user-centric Application Performance Management solution can help inspire your users with every applicati...
It's 2:15pm on a Friday, and I'm sitting in the keynote hall at PyCon 2013 fidgeting through a succession of lightning talks that have very little relevance to my life. Topics like "Python code coverage techniques" (ho-hum) and "Controlling Christmas lights with Python” (yawn - I wonder if there's anything new on Hacker News)...when Solomon Hykes takes the stage, unveils Docker, and the world shifts. If you haven't seen it yet, you should watch the video of Solomon's Pycon The Future of Linux C...
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 and business needs change, data repositories and structures get outmoded rapidly, resulting in increased work for applica...
Enthusiasm for the Internet of Things has reached an all-time high. In 2013 alone, venture capitalists spent more than $1 billion dollars investing in the IoT space. With "smart" appliances and devices, IoT covers wearable smart devices, cloud services to hardware companies. Nest, a Google company, detects temperatures inside homes and automatically adjusts it by tracking its user's habit. These technologies are quickly developing and with it come challenges such as bridging infrastructure gaps,...
NuoDB just introduced the Swifts 2.1 Release. In this demo at 15th Cloud Expo, Seth Proctor, CTO of NuoDB, Inc., discussed why scaling databases in the cloud is challenging, why building your application on top of the infrastructure that is designed with this in mind makes a difference, and what you can do with NuoDB that simplifies your programming model, your operations model.
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...
The 17th International Cloud Expo has announced that its Call for Papers is open. 17th International Cloud Expo, to be held November 3-5, 2015, at the Santa Clara Convention Center in Santa Clara, CA, brings together Cloud Computing, APM, APIs, Microservices, Security, Big Data, Internet of Things, DevOps 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 bu...
As enterprises engage with Big Data technologies to develop applications needed to meet operational demands, new computation fabrics are continually being introduced. To leverage these new innovations, organizations are sacrificing market opportunities to gain expertise in learning new systems. In his session at Big Data Expo, Supreet Oberoi, Vice President of Field Engineering at Concurrent, Inc., discussed how to leverage existing infrastructure and investments and future-proof them against e...
Once the decision has been made to move part or all of a workload to the cloud, a methodology for selecting that workload needs to be established. How do you move to the cloud? What does the discovery, assessment and planning look like? What workloads make sense? Which cloud model makes sense for each workload? What are the considerations for how to select the right cloud model? And how does that fit in with the overall IT transformation?
Cloud Expo New York is happening from June 9 - 11. This event brings together the worlds of Cloud Computing, DevOps, IoT, WebRTC, Big Data and SDDC. We hope to see you there-members of the Blue Box team will exhibit in booth 218 next to the DevOps area. Plus, our Chief Product Officer, Hernan Alvarez, will present his talk "The Cloud Has a Down-and-Dirty Lining" as part of the Operations track in the DevOps Summit portion of the event on June 9 at 11 am. Learn more about his session her...
With major technology companies and startups seriously embracing IoT strategies, now is the perfect time to attend @ThingsExpo in Silicon Valley. Learn what is going on, contribute to the discussions, and ensure that your enterprise is as "IoT-Ready" as it can be! Internet of @ThingsExpo, taking place Nov 3-5, 2015, at the Santa Clara Convention Center in Santa Clara, CA, is co-located with 17th Cloud Expo and will feature technical sessions from a rock star conference faculty and the leading in...
When OpenStack aficionados gather in Vancouver in a couple of weeks, one of the hot topics will be containers, a “new” alternative to virtualization. Actually, container technology has been around for a couple of decades, but it is trending among the IT community at a fever pitch these days and stands to have a huge impact on the future of cloud computing.The appeal of container technology is easy to appreciate. In a nutshell, containers can enable you to run many more applications on the same h...
Docker is an open platform for developers and sysadmins of distributed applications that enables them to build, ship, and run any app anywhere. Docker allows applications to run on any platform irrespective of what tools were used to build it making it easy to distribute, test, and run software. I found this 5 Minute Docker video, which is very helpful when you want to get a quick and digestible overview. If you want to learn more, you can go to Docker’s web page and start with this Docker intro...
The 5th International DevOps Summit, co-located with 17th International Cloud Expo – being held November 3-5, 2015, at the Santa Clara Convention Center in Santa Clara, CA – announces that its Call for Papers is open. Born out of proven success in agile development, cloud computing, and process automation, DevOps is a macro trend you cannot afford to miss. From showcase success stories from early adopters and web-scale businesses, DevOps is expanding to organizations of all sizes, including the...
Over the years, a variety of methodologies have emerged in order to overcome the challenges related to project constraints. The successful use of each methodology seems highly context-dependent. However, communication seems to be the common denominator of the many challenges that project management methodologies intend to resolve. In this respect, Information and Communication Technologies (ICTs) can be viewed as powerful tools for managing projects. Few research papers have focused on the way...