Welcome!

@DXWorldExpo Authors: Yeshim Deniz, Zakia Bouachraoui, Liz McMillan, Pat Romanski, Elizabeth White

Related Topics: @DXWorldExpo, Java IoT, Machine Learning , @CloudExpo, Apache, SDN Journal

@DXWorldExpo: 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.


DXWorldEXPO Digital Transformation Stories
Nicolas Fierro is CEO of MIMIR Blockchain Solutions. He is a programmer, technologist, and operations dev who has worked with Ethereum and blockchain since 2014. His knowledge in blockchain dates to when he performed dev ops services to the Ethereum Foundation as one the privileged few developers to work with the original core team in Switzerland.
DXWorldEXPO LLC announced today that Nutanix has been named "Platinum Sponsor" of CloudEXPO | DevOpsSUMMIT | DXWorldEXPO New York, which will take place November 12-13, 2018 in New York City. Nutanix makes infrastructure invisible, elevating IT to focus on the applications and services that power their business. The Nutanix Enterprise Cloud Platform blends web-scale engineering and consumer-grade design to natively converge server, storage, virtualization and networking into a resilient, softwar...
René Bostic is the Technical VP of the IBM Cloud Unit in North America. Enjoying her career with IBM during the modern millennial technological era, she is an expert in cloud computing, DevOps and emerging cloud technologies such as Blockchain. Her strengths and core competencies include a proven record of accomplishments in consensus building at all levels to assess, plan, and implement enterprise and cloud computing solutions. René is a member of the Society of Women Engineers (SWE) and a m...
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 settlement products to hedge funds and investment banks. After, he co-founded a revenue cycle management company where he learned about Bitcoin and eventually Ethereal. Andrew's role at ConsenSys Enterprise is a mul...
When building large, cloud-based applications that operate at a high scale, it’s important to maintain a high availability and resilience to failures. In order to do that, you must be tolerant of failures, even in light of failures in other areas of your application. “Fly two mistakes high” is an old adage in the radio control airplane hobby. It means, fly high enough so that if you make a mistake, you can continue flying with room to still make mistakes. In his session at 18th Cloud Expo, Lee A...
Digital transformation is about embracing digital technologies into a company's culture to better connect with its customers, automate processes, create better tools, enter new markets, etc. Such a transformation requires continuous orchestration across teams and an environment based on open collaboration and daily experiments. In his session at 21st Cloud Expo, Alex Casalboni, Technical (Cloud) Evangelist at Cloud Academy, explored and discussed the most urgent unsolved challenges to achieve fu...
Daniel Jones is CTO of EngineerBetter, helping enterprises deliver value faster. Previously he was an IT consultant, indie video games developer, head of web development in the finance sector, and an award-winning martial artist. Continuous Delivery makes it possible to exploit findings of cognitive psychology and neuroscience to increase the productivity and happiness of our teams.
Only Adobe gives everyone - from emerging artists to global brands - everything they need to design and deliver exceptional digital experiences. Adobe Systems Incorporated develops, markets, and supports computer software products and technologies. The Company's products allow users to express and use information across all print and electronic media. The Company's Digital Media segment provides tools and solutions that enable individuals, small and medium businesses and enterprises to cre...
Dynatrace is an application performance management software company with products for the information technology departments and digital business owners of medium and large businesses. Building the Future of Monitoring with Artificial Intelligence. Today we can collect lots and lots of performance data. We build beautiful dashboards and even have fancy query languages to access and transform the data. Still performance data is a secret language only a couple of people understand. The more busine...
ICC is a computer systems integrator and server manufacturing company focused on developing products and product appliances to meet a wide range of computational needs for many industries. Their solutions provide benefits across many environments, such as datacenter deployment, HPC, workstations, storage networks and standalone server installations. ICC has been in business for over 23 years and their phenomenal range of clients include multinational corporations, universities, and small busines...