|By Michael Kopp||
|December 12, 2012 07:30 AM EST||
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We recognize that the average no longer represents the bulk of the transactions but can be a lot higher than the median.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!
|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?
Containers and microservices have become topics of intense interest throughout the cloud developer and enterprise IT communities. Accordingly, attendees at the upcoming 16th Cloud Expo at the Javits Center in New York June 9-11 will find fresh new content in a new track called PaaS | Containers & Microservices Containers are not being considered for the first time by the cloud community, but a current era of re-consideration has pushed them to the top of the cloud agenda. With the launch ...
Apr. 26, 2015 07:00 PM EDT Reads: 2,892
While DevOps most critically and famously fosters collaboration, communication, and integration through cultural change, culture is more of an output than an input. In order to actively drive cultural evolution, organizations must make substantial organizational and process changes, and adopt new technologies, to encourage a DevOps culture. Moderated by Andi Mann, panelists will discuss how to balance these three pillars of DevOps, where to focus attention (and resources), where organizations m...
Apr. 26, 2015 05:15 PM EDT Reads: 2,051
of cloud, colocation, managed services and disaster recovery solutions, will exhibit at SYS-CON's 16th International Cloud Expo®, which will take place on June 9-11, 2015, at the Javits Center in New York City, NY. TierPoint, LLC, is a leading national provider of information technology and data center services, including cloud, colocation, disaster recovery and managed IT services, with corporate headquarters in St. Louis, MO. TierPoint was formed through the strategic combination of some of t...
Apr. 26, 2015 04:30 PM EDT Reads: 1,667
SYS-CON Events announced today that Column Technologies, a global technology solutions company, will exhibit at SYS-CON's DevOps Summit 2015 New York, which will take place June 9-11, 2015, at the Javits Center in New York City, NY. Established in 1998, Column Technologies is a leader in application performance and infrastructure management for commercial and federal markets. The company is headquartered in the United States, with a diverse and talented team of more than 350 employees around th...
Apr. 26, 2015 04:00 PM EDT Reads: 1,807
SYS-CON Events announced today that Ciqada will exhibit at SYS-CON's @ThingsExpo, which will take place on June 9-11, 2015, at the Javits Center in New York City, NY. Ciqada™ makes it easy to connect your products to the Internet. By integrating key components - hardware, servers, dashboards, and mobile apps - into an easy-to-use, configurable system, your products can quickly and securely join the internet of things. With remote monitoring, control, and alert messaging capability, you will mee...
Apr. 26, 2015 03:00 PM EDT Reads: 1,934
I woke up this morning to the devastating news about the earthquake in Nepal. Sitting here in California that destruction is literally on the other side of the world but my mind immediately went to thinking about my good friend Jeremy Geelan. See Jeremy and his family have been living in Kathmandu for a while now. His wife, in fact, is the Danish Ambassador to Nepal!
Apr. 26, 2015 02:15 PM EDT Reads: 194
Public Cloud IaaS started it's life in the developer and startup communities and has grown rapidly to a $20B+ industry, but it still pales in comparison to how much is spent worldwide on IT: $3.6 trillion. In fact, there are 8.6 million data centers worldwide, the reality is many small and medium sized business have server closets and colocation footprints filled with servers and storage gear. While on-premise environment virtualization may have peaked at 75%, the Public Cloud has lagged in ado...
Apr. 26, 2015 02:00 PM EDT Reads: 1,322
Dave will share his insights on how Internet of Things for Enterprises are transforming and making more productive and efficient operations and maintenance (O&M) procedures in the cleantech industry and beyond. Speaker Bio: Dave Landa is chief operating officer of Cybozu Corp (kintone US). Based in the San Francisco Bay Area, Dave has been on the forefront of the Cloud revolution driving strategic business development on the executive teams of multiple leading Software as a Services (SaaS) ap...
Apr. 26, 2015 02:00 PM EDT Reads: 1,579
ProfitBricks, the provider of painless cloud infrastructure IaaS, today released its SDK for Ruby, written against the company's new RESTful API. The new SDK joins ProfitBricks' previously announced support for the popular multi-cloud open-source Fog project. This new Ruby SDK, which exposes advanced functionality to take advantage of ProfitBricks' simplicity and productivity, aligns with ProfitBricks' mission to provide a painless way to automate infrastructure in the cloud. Ruby is a genera...
Apr. 26, 2015 01:00 PM EDT Reads: 1,085
This digest provides an overview of good resources that are well worth reading. We’ll be updating this page as new content becomes available, so I suggest you bookmark it. Also, expect more digests to come on different topics that make all of our IT-hearts go boom!
Apr. 26, 2015 01:00 PM EDT Reads: 1,011
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...
Apr. 26, 2015 01:00 PM EDT Reads: 2,240
Microsoft is releasing in the near future Azure Service Fabric as a preview beta. Azure Service Fabric is built to run microservices - a complex application consisting of smaller, interlocked components that enables updating components without disrupting service. Microsoft has used this over the past few years internally for many of its own applications and the new release is for general use, a new product. OSIsoft is an early adopter of this system and run with it to expand into the explo...
Apr. 26, 2015 01:00 PM EDT Reads: 1,297
79% of new products miss their launch date. That was the conclusion of a CGT/Sopheon Survey in which the impact of such market misses were also explored. What it didn't dig into was the reason why so many products and projects miss their launch date. When we start digging into the details with respect to applications, we can find at least one causal factor in the delivery process, specifically that portion which focuses on the actual move into production, from which consumers (internal and...
Apr. 26, 2015 12:45 PM EDT Reads: 949
This is a no-hype, pragmatic post about why I think you should consider architecting your next project the way SOA and/or microservices suggest. No matter if it’s a greenfield approach or if you’re in dire need of refactoring. Please note: considering still keeps open the option of not taking that approach. After reading this, you will have a better idea about whether building multiple small components instead of a single, large component makes sense for your project. This post assumes that you...
Apr. 26, 2015 12:30 PM EDT Reads: 1,163
DevOps tasked with driving success in the cloud need a solution to efficiently leverage multiple clouds while avoiding cloud lock-in. Flexiant today announces the commercial availability of Flexiant Concerto. With Flexiant Concerto, DevOps have cloud freedom to automate the build, deployment and operations of applications consistently across multiple clouds. Concerto is available through four disruptive pricing models aimed to deliver multi-cloud at a price point everyone can afford.
Apr. 26, 2015 12:30 PM EDT Reads: 1,716
This is my first blog post at AppDynamics, and I have to say that it’s great to be aboard. It’s been a hectic first couple of weeks, but the energy, enthusiasm and friendliness of everyone I have met has made me very excited about 2015! AppDynamics has a market leading APM and analytics platform but it also takes great people to make a great company – and AppDynamics has a wealth of talent! So to start my blogging life at AppDynamics I want to focus on something that is a red-hot buzzword in IT...
Apr. 26, 2015 12:15 PM EDT Reads: 1,359
There is no quick way to learn Jython API but to experiment with it. The easiest way is to start with Jytutor extension for XL Deploy. Now you can also use the code snippet for exposing jython/python context in XL Deploy environment by running it directly in Jytutor Here’s how you can go ahead with that Download the Jytutor extension referring to the Jytutor Blog or from the following link https://github.com/xebialabs-community/xld-jytutor-plugin/releases Shutdown your XL Deploy server...
Apr. 26, 2015 12:00 PM EDT Reads: 889
ProfitBricks has launched its new DevOps Central and REST API, along with support for three multi-cloud libraries and a Python SDK. This, combined with its already existing SOAP API and its new RESTful API, moves ProfitBricks into a position to better serve the DevOps community and provide the ability to automate cloud infrastructure in a multi-cloud world. Following this momentum, ProfitBricks has also introduced several libraries that enable developers to use their favorite language to code ...
Apr. 26, 2015 12:00 PM EDT Reads: 1,708
What exactly is a cognitive application? In her session at 16th Cloud Expo, Ashley Hathaway, Product Manager at IBM Watson, will look at the services being offered by the IBM Watson Developer Cloud and what that means for developers and Big Data. She'll explore how IBM Watson and its partnerships will continue to grow and help define what it means to be a cognitive service, as well as take a look at the offerings on Bluemix. She will also check out how Watson and the Alchemy API team up to off...
Apr. 26, 2015 12:00 PM EDT Reads: 1,869
ProfitBricks, the provider of painless cloud infrastructure IaaS, announced the launch of its new DevOps Central and REST API, along with support for three multi-cloud libraries and a Python SDK. This, combined with its already existing SOAP API and its new RESTful API, moves ProfitBricks into a position to better serve the DevOps community and provide the ability to automate cloud infrastructure in a multi-cloud world. Following this momentum, ProfitBricks is also today introducing several l...
Apr. 26, 2015 12:00 PM EDT Reads: 1,201