Welcome!

Java Authors: Liz McMillan, Elizabeth White, Srinivasan Sundara Rajan, Patrick Carey, Pat Romanski

Related Topics: Big Data Journal, Java, AJAX & REA, Cloud Expo, Apache, SDN Journal

Big Data Journal: Article

So What? – Monitoring Hadoop Beyond Ganglia

Don’t just run Hadoop jobs at scale, run them efficiently and at scale

Over the last couple of months I have been talking to more and more customers who are either bringing their Hadoop clusters into production or have already done so and are now getting serious about operations. This leads to some interesting discussions about how to monitor Hadoop properly and one thing pops up quite often: Do they need anything beyond Ganglia? If yes, what should they do beyond it?

The Basics
As in every other system, monitoring in a Hadoop environment starts with the basics: System Metrics - CPU, Disk, Memory you know the drill. Of special importance in a Hadoop system is a well-balanced cluster; you don't want to have some nodes being much more (or less) utilized then others. Besides CPU and memory utilization, Disk utilization and of course I/O throughput is of high importance. After all the most likely bottleneck in a Big Data system is I/O - either with ingress (network and disk), moving data around (e.g., MapReduce shuffle on the network) and straightforward read/write to disk.

The problem in a Hadoop system is of course its size. Nothing new for us, some of our customers monitor well beyond 1000+ JVMs with CompuwareAPM. The "advantage" in a Hadoop system is its relative conformity - every node looks pretty much like the other. This is what Ganglia leverages.

Cluster Monitoring with Ganglia
What Ganglia is very good at is providing an overview over how a cluster is utilized. The load chart is particularly interesting:

This chart shows the CPU load on a 1000 Server cluster that has roughly 15.000 CPUs

It tells us the number of available cores in the system and the number of running processes (in theory a core can never handle more than one process at a time) and the 1-min load average. If the system is getting fully utilized the 1-min load average would approach the total number of CPUs. Another view on this is the well-known CPU utilization chart:

CPU Utilization over the last day. While the utilization stays well below 10% we see a lot of I/O wait spikes.

While the load chart gives a good overall impression of usage, the utilization tells us the story of how the CPUs are used. While typical CPU charts show a single server, Ganglia specializes in showing whole clusters (the picture shows CPU usage of a 1000 machine cluster). In the case of the depicted chart we see that the CPUs are experiencing a lot of I/O wait spikes, which points toward heavy disk I/O. Basically it seems the disk I/O is the reason that we cannot utilize our CPU better at these times. But in general our cluster is well underutilized in terms of CPU.

Trends are also easy to understand, as can be seen in this memory chart over a year.

Memory capacity and usage over a year

All this looks pretty good, so what is missing? The "so what" and "why" is what is missing. If my memory demand is growing, I have no way of knowing why it is growing. If the CPU chart tells me that I spend a lot of time waiting, it does not tell what to do, or why that is so? These questions are beyond the scope of Ganglia.

What about Hadoop specifics?
Ganglia also has a Hadoop plugin, which basically gives you access to all the usual Hadoop metrics (unfortunately a comprehensive list of Hadoop metrics is really hard to find, appreciate if somebody commented the link). There is a good explanation on what is interesting on Edward Caproli's page: JoinTheGrid. Basically you can use those metrics to monitor the capacity and usage trends of HDFS and the NameNodes and also how many jobs, mappers and reducers are running.

Capacity of the DataNodes over time

Capacity of the Name Nodes over time

The DataNode Operations give me an impression of I/O pressure on the Hadoop cluster

All these charts can of course be easily built in any modern monitoring or APM solution like CompuwareAPM, but Ganglia gives you a simple starting point; and it's Free as in Beer.

What's missing again is the so what? If my jobs are running a lot longer than yesterday, what should I do? Why do they run longer? A Hadoop expert might dig into 10 different charts around I/O and Network, spilling, look at log files among other things and try an educated guess as to what might be the problem. But we aren't all experts, neither do we have the time to dig into all of these metrics and log files all the time.

This is the reason that we and our customers are moving beyond Ganglia - to solve the "Why" and "So What" within time constraints.

Beyond the Basics #1 - Understanding Cluster Utilization
A use case that we get from customers is that they want to know which users or which pools (in case of the fair scheduler) are responsible for how much of the cluster utilization. LinkedIn just released White Elephant, a tool that parses MapReduce logs and builds some nice dashboards and shows you which of your users occupy how much of your cluster. This is of course based on log file analysis and thus okay for analysis but not for monitoring. With proper tools in place we can do the same thing in near real time.

The CPU Usage in the Hadoop Cluster on per User basis

In this example I wanted to monitor which user consumed how much of my Amazon EMR cluster. If we see a user or pool that occupies a lot of the cluster we can course also see which jobs are running and how much of the cluster they occupy.

The CPU Usage in the Hadoop Cluster on per Job basis

And this will also tell us if that job has always been there, and just uses a lot more resources now. This would be our cue to start analyzing what has changed.

Beyond the Basics #2 - Understanding why my jobs are slow(er)
If we want to understand why a job is slow we need to look at a high-level break down first.

In which phase of the map reduce do we spend the most time, or did we spend more time than yesterday? Understanding these timings in context with the respective job counters, like Map Input or Spilled Records, helps us understand why the phase took longer.

Overview of the time spent in different phases and the respective input/output counters

At this point we will already have a pretty good idea as to what happened. We either simply have more data to crunch (more input data) or a portion of the MapReduce job consumes more CPU (code change?) or we spill more records to disk (code change or Hadoop config change?). We might also detect an unbalanced cluster in the performance breakdown.

This job is executing nearly exclusively on a single node instead of distributing

In this case we want to check whether all the involved nodes processed the same amount of data

Here we see that there is a wide range from minimum, average to maximum on mapped input and output records. The data is not balanced

or if the difference can again be found in the code (different kinds of computations). If we are running against HBase we might of course have an issue with HBase performance or distribution.

At the beginning of the job only a single HBase region Server consumes CPU while all others remain idle

On the other hand, if a lot of mapping time is spent in the garbage collector then you should maybe invest in larger JVMs.

The Performance Breakdown of this particular job shows considerable time in GC suspension

If spilling data to disk is where we spend our time, we should take a closer look at that phase. It might turn out that all of our time is spent on disk wait.

If the Disk were the bottleneck we would see it on disk I/O here

Now if disk write is our bottleneck, then really the only thing that we can do is reduce the map output records. Adding a combiner will not reduce the disk write (it will actually increase it, read here). In other words combining only optimizes the shuffle phase, thus the amount of data sent over the network, but not spill time!!

And at the very detailed level we can look at single task executions and understand in detail what is really going on.

The detailed data about each Map, Reduce Task Atttempt as well as the spills and shuffles

Conclusion
Ganglia is a great tool for high-level monitoring of your Hadoop cluster utilization, but it is not enough. The fact that everybody is working on additional means to understand the Hadoop cluster (Hortonworks with Ambari, Cloudera with their Manager, LinkedIn with White Elephant, the Star Fish project...) shows that there is a lot more needed beyond simple monitoring. Even those more advanced monitoring tools are not always answering the "why" though, which is what we really need to do. This is where the Performance Management discipline can add a lot of value and really help you get the best out of your Hadoop cluster. In other words don't just run Hadoop jobs at scale, run them efficiently and at scale!

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.


@ThingsExpo Stories
“In the past year we've seen a lot of stabilization of WebRTC. You can now use it in production with a far greater degree of certainty. A lot of the real developments in the past year have been in things like the data channel, which will enable a whole new type of application," explained Peter Dunkley, Technical Director at Acision, in this SYS-CON.tv interview at @ThingsExpo, held Nov 4–6, 2014, at the Santa Clara Convention Center in Santa Clara, CA.
SYS-CON Events announced today that Windstream, a leading provider of advanced network and cloud communications, has been named “Silver Sponsor” of SYS-CON's 16th International Cloud Expo®, which will take place on June 9–11, 2015, at the Javits Center in New York, NY. Windstream (Nasdaq: WIN), a FORTUNE 500 and S&P 500 company, is a leading provider of advanced network communications, including cloud computing and managed services, to businesses nationwide. The company also offers broadband, phone and digital TV services to consumers primarily in rural areas.
The Internet of Things is not new. Historically, smart businesses have used its basic concept of leveraging data to drive better decision making and have capitalized on those insights to realize additional revenue opportunities. So, what has changed to make the Internet of Things one of the hottest topics in tech? In his session at @ThingsExpo, Chris Gray, Director, Embedded and Internet of Things, discussed the underlying factors that are driving the economics of intelligent systems. Discover how hardware commoditization, the ubiquitous nature of connectivity, and the emergence of Big Data a...

ARMONK, N.Y., Nov. 20, 2014 /PRNewswire/ --  IBM (NYSE: IBM) today announced that it is bringing a greater level of control, security and flexibility to cloud-based application development and delivery with a single-tenant version of Bluemix, IBM's platform-as-a-service. The new platform enables developers to build ap...

SYS-CON Events announced today that IDenticard 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. IDenticard™ is the security division of Brady Corp (NYSE: BRC), a $1.5 billion manufacturer of identification products. We have small-company values with the strength and stability of a major corporation. IDenticard offers local sales, support and service to our customers across the United States and Canada. Our partner network encompasses some 300 of the world's leading systems integrators and security s...
DevOps Summit 2015 New York, co-located with the 16th International Cloud Expo - to be held June 9-11, 2015, at the Javits Center in New York City, NY - announces that it is now accepting Keynote Proposals. 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 wait for long development cycles that produce software that is obsolete at launch. DevOps may be disruptive, but it is essential.
The major cloud platforms defy a simple, side-by-side analysis. Each of the major IaaS public-cloud platforms offers their own unique strengths and functionality. Options for on-site private cloud are diverse as well, and must be designed and deployed while taking existing legacy architecture and infrastructure into account. Then the reality is that most enterprises are embarking on a hybrid cloud strategy and programs. In this Power Panel at 15th Cloud Expo (http://www.CloudComputingExpo.com), moderated by Ashar Baig, Research Director, Cloud, at Gigaom Research, Nate Gordon, Director of T...
"BSQUARE is in the business of selling software solutions for smart connected devices. It's obvious that IoT has moved from being a technology to being a fundamental part of business, and in the last 18 months people have said let's figure out how to do it and let's put some focus on it, " explained Dave Wagstaff, VP & Chief Architect, at BSQUARE Corporation, in this SYS-CON.tv interview at @ThingsExpo, held Nov 4-6, 2014, at the Santa Clara Convention Center in Santa Clara, CA.
"People are a lot more knowledgeable about APIs now. There are two types of people who work with APIs - IT people who want to use APIs for something internal and the product managers who want to do something outside APIs for people to connect to them," explained Roberto Medrano, Executive Vice President at SOA Software, in this SYS-CON.tv interview at Cloud Expo, held Nov 4–6, 2014, at the Santa Clara Convention Center in Santa Clara, CA.
Nigeria has the largest economy in Africa, at more than US$500 billion, and ranks 23rd in the world. A recent re-evaluation of Nigeria's true economic size doubled the previous estimate, and brought it well ahead of South Africa, which is a member (unlike Nigeria) of the G20 club for political as well as economic reasons. Nigeria's economy can be said to be quite diverse from one point of view, but heavily dependent on oil and gas at the same time. Oil and natural gas account for about 15% of Nigera's overall economy, but traditionally represent more than 90% of the country's exports and as...
The Internet of Things is a misnomer. That implies that everything is on the Internet, and that simply should not be - especially for things that are blurring the line between medical devices that stimulate like a pacemaker and quantified self-sensors like a pedometer or pulse tracker. The mesh of things that we manage must be segmented into zones of trust for sensing data, transmitting data, receiving command and control administrative changes, and peer-to-peer mesh messaging. In his session at @ThingsExpo, Ryan Bagnulo, Solution Architect / Software Engineer at SOA Software, focused on desi...
"At our booth we are showing how to provide trust in the Internet of Things. Trust is where everything starts to become secure and trustworthy. Now with the scaling of the Internet of Things it becomes an interesting question – I've heard numbers from 200 billion devices next year up to a trillion in the next 10 to 15 years," explained Johannes Lintzen, Vice President of Sales at Utimaco, in this SYS-CON.tv interview at @ThingsExpo, held Nov 4–6, 2014, at the Santa Clara Convention Center in Santa Clara, CA.
"For over 25 years we have been working with a lot of enterprise customers and we have seen how companies create applications. And now that we have moved to cloud computing, mobile, social and the Internet of Things, we see that the market needs a new way of creating applications," stated Jesse Shiah, CEO, President and Co-Founder of AgilePoint Inc., in this SYS-CON.tv interview at 15th Cloud Expo, held Nov 4–6, 2014, at the Santa Clara Convention Center in Santa Clara, CA.
SYS-CON Events announced today that Gridstore™, the leader in hyper-converged infrastructure purpose-built to optimize Microsoft workloads, 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. Gridstore™ is the leader in hyper-converged infrastructure purpose-built for Microsoft workloads and designed to accelerate applications in virtualized environments. Gridstore’s hyper-converged infrastructure is the industry’s first all flash version of HyperConverged Appliances that include both compute and storag...
Today’s enterprise is being driven by disruptive competitive and human capital requirements to provide enterprise application access through not only desktops, but also mobile devices. To retrofit existing programs across all these devices using traditional programming methods is very costly and time consuming – often prohibitively so. In his session at @ThingsExpo, Jesse Shiah, CEO, President, and Co-Founder of AgilePoint Inc., discussed how you can create applications that run on all mobile devices as well as laptops and desktops using a visual drag-and-drop application – and eForms-buildi...
We certainly live in interesting technological times. And no more interesting than the current competing IoT standards for connectivity. Various standards bodies, approaches, and ecosystems are vying for mindshare and positioning for a competitive edge. It is clear that when the dust settles, we will have new protocols, evolved protocols, that will change the way we interact with devices and infrastructure. We will also have evolved web protocols, like HTTP/2, that will be changing the very core of our infrastructures. At the same time, we have old approaches made new again like micro-services...
Code Halos - aka "digital fingerprints" - are the key organizing principle to understand a) how dumb things become smart and b) how to monetize this dynamic. In his session at @ThingsExpo, Robert Brown, AVP, Center for the Future of Work at Cognizant Technology Solutions, outlined research, analysis and recommendations from his recently published book on this phenomena on the way leading edge organizations like GE and Disney are unlocking the Internet of Things opportunity and what steps your organization should be taking to position itself for the next platform of digital competition.
The 3rd International Internet of @ThingsExpo, co-located with the 16th International Cloud Expo - to be held June 9-11, 2015, at the Javits Center in New York City, NY - announces that its Call for Papers is now open. The Internet of Things (IoT) is the biggest idea since the creation of the Worldwide Web more than 20 years ago.
As the Internet of Things unfolds, mobile and wearable devices are blurring the line between physical and digital, integrating ever more closely with our interests, our routines, our daily lives. Contextual computing and smart, sensor-equipped spaces bring the potential to walk through a world that recognizes us and responds accordingly. We become continuous transmitters and receivers of data. In his session at @ThingsExpo, Andrew Bolwell, Director of Innovation for HP's Printing and Personal Systems Group, discussed how key attributes of mobile technology – touch input, sensors, social, and ...
In their session at @ThingsExpo, Shyam Varan Nath, Principal Architect at GE, and Ibrahim Gokcen, who leads GE's advanced IoT analytics, focused on the Internet of Things / Industrial Internet and how to make it operational for business end-users. Learn about the challenges posed by machine and sensor data and how to marry it with enterprise data. They also discussed the tips and tricks to provide the Industrial Internet as an end-user consumable service using Big Data Analytics and Industrial Cloud.