Managing the File Explosion

By Gary Orenstein

August 26, 2008

The “digital universe” seems to be mirroring our own physical universe: it just keeps expanding at faster and faster rates. Consider this jaw-dropping statistic from market researchers at IDC: the total volume of digital information that is created and replicated globally reached 281 BILLION gigabytes of digital information in 2007. That’s 45 gigabytes for every person on Earth.

A big part of this growth results from the rise in consumer-generated content. While the initial stages of the Web included a relatively small number of sites with relatively fixed content, we now have thousands of choices. And many Web destinations are driven by their ability to aggregate an enormous number of consumer-generated files such as photos and videos.Figure 1: The File Explosion
But there is more to the story. Even though the amount of data being created and stored continues to rise at a significant clip, the number of files, or file count, is rising at a much faster rate. Recent data from IDC details how the compound annual growth rate for terabytes of capacity is 59 percent while the compound growth rate for file count is 88 percent during the same period. This results in new challenges and demands for Web-scale infrastructure.

No longer is finding enough capacity sufficient to scale Web-scale infrastructure for hundreds of thousands to millions of users. The thornier problem is now managing the “file count explosion” resulting from soaring numbers of individually-contributed user-generated content items and optimizing the performance of increasingly-complex Web-scale applications across Web servers such as Apache and Lighttpd, indexing engines such as Lucene, and databases such as MySQL.

On the file count side of the ledger, it’s a classic “needle in the haystack” problem. As more and more files are generated, it takes longer and longer to handle user queries for specific files. Why? The indexes used to find files are growing bigger as well. Keep in mind that in most of these applications, even though the content requested varies widely across each individual user, the ways to reach that content such as searching by name, subject, or “recency” requires the use of a common index for all users. As a result, it takes more time to comb through these enormous indexes to find the files a user wants. And this creates serious performance bottlenecks when the historical practice of keeping that indexing information on mechanical disks starts to bog down.

The root of this index overload is metadata, or the information about files, rather than the file content itself. Specifically, when file systems need to search through lots of files and numerous directory levels, the overhead of those metadata operations can overwhelm the storage system. The impact of this can be seen in a typical file or object retrieval operation where the common request to “walk the directory tree” leads to an excessive number of additional NFS operations instead of simply being able to identify the file location and getting it.

Solving this bottleneck requires a reevaluation of how to deploy memory in the datacenter. Forward-looking datacenter architects and managers are now recognizing the advantages of scalable, network-based caching as a means of solving this growing “file explosion” problem. By making a scalable pool of memory available in the form of a centralized network caching service, datacenters can deliver accelerated file services that offload slower mechanical disk drives, storage subsystems and existing file systems from the heavy lifting of handling this mushrooming metadata and index growth. Some advantages of network caching include:

1. Peak performance

Deploying memory as a centralized cache enhances disk drives performance and enables immediate I/O performance gains of 10-50X, significantly improving application response times.

2. Consolidated and centralized resources

Implementing a consolidated, shared network caching resource dramatically improves overall system efficiency, reduces overall costs, and intelligently rebalances workloads in response to existing application infrastructure.

3. Single-point management

A centralized model of single-point management allows for true expansion without additional management overhead. Adding resources to the caching appliances does not require extra administrative tasks as a single appliance can scale under one management interface.

4. Use of existing infrastructure

Centralized caching solutions enhance existing infrastructure by supporting current storage systems and Web/application servers.

5. Multi-use deployment

As a shared resource, scalable caching appliances can optimize the performance of multiple applications within a single appliance, greatly increasing the overall efficiency and effectiveness in the datacenter.

To address Web-scale application performance issues, new accelerated file and object services are emerging to remedy the scaling challenges. These new services make applications available to more concurrent users with greatly improved file access performance to meet the compute and storage intensive demands of Web-scale datacenters.

One example of these services includes the ability to handle and cache metadata selectively. The conventional model of storing all information on disk-based systems can lead to lengthy response times, limiting the number of application transactions that can take place within a given time frame. Metadata operations such as looking up file names and locations can involve numerous small, random requests that — when multiplied across thousands of simultaneous users — often overwhelm a typical disk storage system. When the file count grows, and the number of nested directories increases, it is not impossible to have ratios up to 20:1 of metadata commands to file retrieval commands. This overhead limits the scale of the overall infrastructure.

By keeping frequently requested data in memory as opposed to on mechanical disks, system I/O performance increases dramatically, enabling improved application response. The primary benefit of the centralized caching deployment is the ability to optimize requests across hundreds or thousands of Web servers. This in turn enables Web applications to easily scale the number of concurrent users without suffering from excessive delays or the need to grossly overprovision infrastructure.

Conclusions

Application workloads are changing from individual users accessing desktop applications to hundreds or thousands of concurrent users accessing common Web applications. Whether for social networking, file sharing, or commercial applications, the use of consolidated Web-scale infrastructure is on the rise, demanding new thinking for datacenter systems.

Fortunately, not everything has to be re-architected. As the demands for rapid file access increase, placing heavier and heavier loads on original Web infrastructure, Web-scale applications can make use of centralized caching resources to scale appropriately with existing infrastructure. This ensures that rapid response times are met for an optimized end user experience.

About the Author

Gary Orenstein is vice president of marketing and business development for Gear6 and has been active in the IP storage networking industry since its inception. He was an initial governing board member of the Storage Networking Industry Association IP Storage Forum where he helped develop, promote, and deliver educational information furthering market growth.

Prior to Gear6, Gary served as vice president of marketing at Compellent Technologies, a network storage company delivering affordable, modular SANs. Before Compellent, he was a founding team member at Nishan Systems, a leader in the IP Storage market, and spearheaded many of Nishan’s milestone industry events such as The Promontory Project – the first transcontinental IP storage network, and the first wire-speed iSCSI demonstration. Prior to that, he spent several years building international distribution and joint ventures for US companies, including a distributorship for Sun Microsystems in Asia.

Gary is also the author of IP Storage Networking: Straight to the Core — a book that outlines the business value of enterprise storage technology. He holds an MBA from the Wharton School at the University of Pennsylvania, and a BA from Dartmouth College.

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion XL — were added to the benchmark suite as MLPerf continues Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing power it brings to artificial intelligence.  Nvidia's DGX Read more…

Call for Participation in Workshop on Potential NSF CISE Quantum Initiative

March 26, 2024

Editor’s Note: Next month there will be a workshop to discuss what a quantum initiative led by NSF’s Computer, Information Science and Engineering (CISE) directorate could entail. The details are posted below in a Ca Read more…

Waseda U. Researchers Reports New Quantum Algorithm for Speeding Optimization

March 25, 2024

Optimization problems cover a wide range of applications and are often cited as good candidates for quantum computing. However, the execution time for constrained combinatorial optimization applications on quantum device Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at the network layer threatens to make bigger and brawnier pro Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HBM3E memory as well as the the ability to train 1 trillion pa Read more…

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HB Read more…

Nvidia Looks to Accelerate GenAI Adoption with NIM

March 19, 2024

Today at the GPU Technology Conference, Nvidia launched a new offering aimed at helping customers quickly deploy their generative AI applications in a secure, s Read more…

The Generative AI Future Is Now, Nvidia’s Huang Says

March 19, 2024

We are in the early days of a transformative shift in how business gets done thanks to the advent of generative AI, according to Nvidia CEO and cofounder Jensen Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, codenamed Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. Wh Read more…

Alibaba Shuts Down its Quantum Computing Effort

November 30, 2023

In case you missed it, China’s e-commerce giant Alibaba has shut down its quantum computing research effort. It’s not entirely clear what drove the change. Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Shutterstock 1285747942

AMD’s Horsepower-packed MI300X GPU Beats Nvidia’s Upcoming H200

December 7, 2023

AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…

DoD Takes a Long View of Quantum Computing

December 19, 2023

Given the large sums tied to expensive weapon systems – think $100-million-plus per F-35 fighter – it’s easy to forget the U.S. Department of Defense is a Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

Intel’s Server and PC Chip Development Will Blur After 2025

January 15, 2024

Intel's dealing with much more than chip rivals breathing down its neck; it is simultaneously integrating a bevy of new technologies such as chiplets, artificia Read more…

Baidu Exits Quantum, Closely Following Alibaba’s Earlier Move

January 5, 2024

Reuters reported this week that Baidu, China’s giant e-commerce and services provider, is exiting the quantum computing development arena. Reuters reported � Read more…

Leading Solution Providers

Contributors

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

Shutterstock 1179408610

Google Addresses the Mysteries of Its Hypercomputer 

December 28, 2023

When Google launched its Hypercomputer earlier this month (December 2023), the first reaction was, "Say what?" It turns out that the Hypercomputer is Google's t Read more…

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

Shutterstock 1606064203

Meta’s Zuckerberg Puts Its AI Future in the Hands of 600,000 GPUs

January 25, 2024

In under two minutes, Meta's CEO, Mark Zuckerberg, laid out the company's AI plans, which included a plan to build an artificial intelligence system with the eq Read more…

Google Introduces ‘Hypercomputer’ to Its AI Infrastructure

December 11, 2023

Google ran out of monikers to describe its new AI system released on December 7. Supercomputer perhaps wasn't an apt description, so it settled on Hypercomputer Read more…

China Is All In on a RISC-V Future

January 8, 2024

The state of RISC-V in China was discussed in a recent report released by the Jamestown Foundation, a Washington, D.C.-based think tank. The report, entitled "E Read more…

Intel Won’t Have a Xeon Max Chip with New Emerald Rapids CPU

December 14, 2023

As expected, Intel officially announced its 5th generation Xeon server chips codenamed Emerald Rapids at an event in New York City, where the focus was really o Read more…

IBM Quantum Summit: Two New QPUs, Upgraded Qiskit, 10-year Roadmap and More

December 4, 2023

IBM kicks off its annual Quantum Summit today and will announce a broad range of advances including its much-anticipated 1121-qubit Condor QPU, a smaller 133-qu Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire