Welcome!

Microservices Expo Authors: Liz McMillan, VictorOps Blog, Pat Romanski, Derek Weeks, AppDynamics Blog

Related Topics: Containers Expo Blog, Java IoT, Microservices Expo, Linux Containers, Open Source Cloud, SDN Journal

Containers Expo Blog: Article

Preventing Performance Bottlenecks with Inline Deduplication

In a typical enterprise storage system the bottleneck to performance is in media bandwidth or computational overhead

Implementing high performance in enterprise storage is a constant battle to find and eliminate the next system bottleneck. Normally this alternates between limits of the underlying media and the computational overhead of metadata management, but choosing the wrong approach to deduplication can introduce a third performance challenge that can be impossible to overcome. Storage that implements a multi-pass approach to data optimization, such as staged or post-process deduplication, becomes inherently at a disadvantage for both computational and media overhead.

Common Performance Bottlenecks
In a typical enterprise storage system the bottleneck to performance is in one of two places: media bandwidth or computational overhead.

For the storage system designer, media overhead is the simplest to address - add more or faster media. In a hard disk-based storage system, this means adding faster drives, more drives, and larger drive sets. In a flash storage system bandwidth is increased by using SLC flash, adding more independent modules, allocating more over-provisioned space, and improving the flash translation layer.

Reducing computational overhead is a greater challenge. Once you have enough media bandwidth the problem becomes shuffling data to and from the storage initiators. Identifying the bottlenecks to performance here can be devilishly complex, as the designer must be concerned about matters such as system memory bandwidth, number of data copies and, especially in today's multi-core world, synchronization between multiple requests. There's no silver bullet here, so the only solution is having a very talented team of software engineers designing and optimizing the storage platform.

Adding deduplication introduces complexity directly into this most challenging area for performance improvement. Any deduplication implementation must interact directly with the storage metadata that is so critical to performance, since I/O requests are being redirected or eliminated based on the system's knowledge of duplicate data. Unless the deduplication technology has been designed and implemented in an inline, multi-core scalable, and low memory overhead way, system architects often try to separate deduplication into a separate layer and a second pass through the data. This is a mistake that harms storage performance in a way that cannot be repaired.

The Impossible Challenge of Multi-Pass Deduplication
This second pass through the data commonly occurs in two possible places: on the final storage media or when transferring data from a staging area to the final storage media.

The first case is always called post-process deduplication. Data are written to their resting media location and a separate process later reads them back, as time and bandwidth allow, determining if any portions are duplicates. If there are duplicates then storage metadata is updated to note this and space is freed for reuse. I've written extensively in the past about the risks of post-process deduplication; since it always requires additional media bandwidth and computational overhead it severely harms performance, and since there are no guarantees about when deduplication will occur it does not meet the requirements for high-change-rate use cases such as VDI.

Post-process Deduplication

The second case, where data is deduplicated as it is being moved from a staging location to a final media location, is often erroneously called inline - as it is inline with that destaging process - but is really just a modified form of post-process deduplication. As with conventional post-process deduplication, another round of data read and processing must occur. Additionally, now both the staging media and final media must provide the full system level of performance or either can become the bottleneck.

For example, some flash storage systems stage all data to a small arena of SLC flash prior to deduplication. This design doubles the number of possible performance bottlenecks in the architecture: performance writing to the staging area, front-end data ingestion, final media storage performance, and the deduplication and de-staging process itself. This sort of multi-pass deduplication process retains all of the negative performance aspects of a traditional post-process implementation.

Staged Post-process Deduplication

High Performance Requires Inline Deduplication
Any form of multi-pass deduplication introduces new bottlenecks that prevent an enterprise storage system from delivering the highest levels of performance. Post-process deduplication, whether on the final media or during a destaging process, creates additional overhead in both media access and data processing. For flash storage platforms requiring the highest levels of performance, only tightly integrated inline deduplication can meet all system requirements.

More Stories By Jered Floyd

Jered Floyd, Chief Technology Officer and Founder of Permabit Technology Corporation, is responsible for exploring strategic future directions for Permabit’s products, and providing thought leadership to guide the company’s data optimization initiatives. He has previously deployed Permabit’s effective software development methodologies and was responsible for developing Permabit product’s core protocol and initial server and system architectures.

Prior to Permabit, Floyd was a Research Scientist on the Microbial Engineering project at the MIT Artificial Intelligence Laboratory, working to bridge the gap between biological and computational systems. Earlier at Turbine, he developed a robust integration language for managing active objects in a massively distributed online virtual environment. Floyd holds Bachelor’s and Master’s degrees in Electrical Engineering and Computer Science from the Massachusetts Institute of Technology.

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.


@MicroservicesExpo Stories
In the world of DevOps there are ‘known good practices’ – aka ‘patterns’ – and ‘known bad practices’ – aka ‘anti-patterns.' Many of these patterns and anti-patterns have been developed from real world experience, especially by the early adopters of DevOps theory; but many are more feasible in theory than in practice, especially for more recent entrants to the DevOps scene. In this power panel at @DevOpsSummit at 18th Cloud Expo, moderated by DevOps Conference Chair Andi Mann, panelists will dis...
Last week I had the pleasure of speaking on a panel at Sapphire Ventures Next-Gen Tech Stack Forum in San Francisco. Obviously, I was excited to join the discussion, but as a participant the event crystallized not only where the larger software development market is relative to microservices, container technologies (like Docker), continuous integration and deployment; but also provided insight into where DevOps is heading in the coming years.
Much of the value of DevOps comes from a (renewed) focus on measurement, sharing, and continuous feedback loops. In increasingly complex DevOps workflows and environments, and especially in larger, regulated, or more crystallized organizations, these core concepts become even more critical. In his session at @DevOpsSummit at 18th Cloud Expo, Andi Mann, Chief Technology Advocate at Splunk, will show how, by focusing on 'metrics that matter,' you can provide objective, transparent, and meaningfu...
Wow, if you ever wanted to learn about Rugged DevOps (some call it DevSecOps), sit down for a spell with Shannon Lietz, Ian Allison and Scott Kennedy from Intuit. We discussed a number of important topics including internal war games, culture hacking, gamification of Rugged DevOps and starting as a small team. There are 100 gold nuggets in this conversation for novices and experts alike.
In a crowded world of popular computer languages, platforms and ecosystems, Node.js is one of the hottest. According to w3techs.com, Node.js usage has gone up 241 percent in the last year alone. Retailers have taken notice and are implementing it on many levels. I am going to share the basics of Node.js, and discuss why retailers are using it to reduce page load times and improve server efficiency. I’ll talk about similar developments such as Docker and microservices, and look at several compani...
The notion of customer journeys, of course, are central to the digital marketer’s playbook. Clearly, enterprises should focus their digital efforts on such journeys, as they represent customer interactions over time. But making customer journeys the centerpiece of the enterprise architecture, however, leaves more questions than answers. The challenge arises when EAs consider the context of the customer journey in the overall architecture as well as the architectural elements that make up each...
Much of the discussion around cloud DevOps focuses on the speed with which companies need to get new code into production. This focus is important – because in an increasingly digital marketplace, new code enables new value propositions. New code is also often essential for maintaining competitive parity with market innovators. But new code doesn’t just have to deliver the functionality the business requires. It also has to behave well because the behavior of code in the cloud affects performan...
Admittedly, two years ago I was a bulk contributor to the DevOps noise with conversations rooted in the movement around culture, principles, and goals. And while all of these elements of DevOps environments are important, I’ve found that the biggest challenge now is a lack of understanding as to why DevOps is beneficial. It’s getting the wheels going, or just taking the next step. The best way to start on the road to change is to take a look at the companies that have already made great headway ...
In 2006, Martin Fowler posted his now famous essay on Continuous Integration. Looking back, what seemed revolutionary, radical or just plain crazy is now common, pedestrian and "just what you do." I love it. Back then, building and releasing software was a real pain. Integration was something you did at the end, after code complete, and we didn't know how long it would take. Some people may recall how we, as an industry, spent a massive amount of time integrating code from one team with another...
Many private cloud projects were built to deliver self-service access to development and test resources. While those clouds delivered faster access to resources, they lacked visibility, control and security needed for production deployments. In their session at 18th Cloud Expo, Steve Anderson, Product Manager at BMC Software, and Rick Lefort, Principal Technical Marketing Consultant at BMC Software, will discuss how a cloud designed for production operations not only helps accelerate developer...
I have an article in the recently released “DZone Guide to Building and Deploying Applications on the Cloud” entitled “Fullstack Engineering in the Age of Hybrid Cloud”. In this article I discuss the need and skills of a Fullstack Engineer with relation to troubleshooting and repairing complex, distributed hybrid cloud applications. My recent experiences with troubleshooting issues with my Docker WordPress container only reinforce the details I wrote about in this piece. Without my comprehensive...
From the conception of Docker containers to the unfolding microservices revolution we see today, here is a brief history of what I like to call 'containerology'. In 2013, we were solidly in the monolithic application era. I had noticed that a growing amount of effort was going into deploying and configuring applications. As applications had grown in complexity and interdependency over the years, the effort to install and configure them was becoming significant. But the road did not end with a ...
Struggling to keep up with increasing application demand? Learn how Platform as a Service (PaaS) can streamline application development processes and make resource management easy.
As the software delivery industry continues to evolve and mature, the challenge of managing the growing list of the tools and processes becomes more daunting every day. Today, Application Lifecycle Management (ALM) platforms are proving most valuable by providing the governance, management and coordination for every stage of development, deployment and release. Recently, I spoke with Madison Moore at SD Times about the changing market and where ALM is headed.
The goal of any tech business worth its salt is to provide the best product or service to its clients in the most efficient and cost-effective way possible. This is just as true in the development of software products as it is in other product design services. Microservices, an app architecture style that leans mostly on independent, self-contained programs, are quickly becoming the new norm, so to speak. With this change comes a declining reliance on older SOAs like COBRA, a push toward more s...
Small teams are more effective. The general agreement is that anything from 5 to 12 is the 'right' small. But of course small teams will also have 'small' throughput - relatively speaking. So if your demand is X and the throughput of a small team is X/10, you probably need 10 teams to meet that demand. But more teams also mean more effort to coordinate and align their efforts in the same direction. So, the challenge is how to harness the power of small teams and yet orchestrate multiples of them...
SYS-CON Events announced today that Peak 10, Inc., a national IT infrastructure and cloud services provider, will exhibit at SYS-CON's 18th International Cloud Expo®, which will take place on June 7-9, 2016, at the Javits Center in New York City, NY. Peak 10 provides reliable, tailored data center and network services, cloud and managed services. Its solutions are designed to scale and adapt to customers’ changing business needs, enabling them to lower costs, improve performance and focus inter...
If there is anything we have learned by now, is that every business paves their own unique path for releasing software- every pipeline, implementation and practices are a bit different, and DevOps comes in all shapes and sizes. Software delivery practices are often comprised of set of several complementing (or even competing) methodologies – such as leveraging Agile, DevOps and even a mix of ITIL, to create the combination that’s most suitable for your organization and that maximize your busines...
Digital means customer preferences and behavior are driving enterprise technology decisions to be sure, but let’s not forget our employees. After all, when we say customer, we mean customer writ large, including partners, supply chain participants, and yes, those salaried denizens whose daily labor forms the cornerstone of the enterprise. While your customers bask in the warm rays of your digital efforts, are your employees toiling away in the dark recesses of your enterprise, pecking data into...
You deployed your app with the Bluemix PaaS and it's gaining some serious traction, so it's time to make some tweaks. Did you design your application in a way that it can scale in the cloud? Were you even thinking about the cloud when you built the app? If not, chances are your app is going to break. Check out this webcast to learn various techniques for designing applications that will scale successfully in Bluemix, for the confidence you need to take your apps to the next level and beyond.