Less is More when Tackling Datacenter Woes

By By Derrick Harris, Editor

May 26, 2008

The problems: low server utilization and labor- and cost-intensive support of applications. The solution: virtualization … or grid computing … or maybe datacenter automation. Well, it’s time to throw another approach into the mix — one that could carve out a large following if it proves as effective as it is unique.

In the traditional application deployment approach, companies must exert lots of effort building a “golden image,” making sure all the scripts, drivers, libraries, etc., are correct. The end result often is a multiple-gigabyte software stack that took days to weeks to perfect and, after testing, takes additional hours to deploy (likely on its own server, which explains the 5-10 percent average utilization rate). “Once you do that, you sort of lock it in time,” says Jerry McLeod, vice president of marketing and business development for FastScale Technology. “You don’t really want to mess with it because you could stop the application from running.”

FastScale has addressed these issues with its Composer suite of products, which McLeod claims allows users to build lightweight application environments in minutes and provision and deploy them in seconds — without worrying about golden images and gigabytes worth of software components. This is accomplished by putting all required software (including the CDs buried under papers in desk drawers) into a component repository, which breaks down the included files into the smallest-possible components. Composer then examines each file, figures out how it works, and places them into a database sorted by function (e.g., kernels, drivers, etc.). According to McLeod, FastScale’s repository turns Red Hat’s roughly 1,500 files into about 330,000 individual components. Once the accompanying software is in, users upload their applications into the repository. Composer then is able to discover all the hardware requirements for each machine in the datacenter, and through the process of application blueprinting finds the links between the application and the OS.

However, McLeod is quick to point out that nothing is built yet. When users want to deploy an application on a specific machine, they simply launch a Dynamic Application Bundle (DAB), which builds the entire software stack on demand and specific to that machine. Thanks to the componentization process, each DAB uses only the necessary pieces, leading to an application image about 99 percent smaller than a golden image. For example, cites McLeod, a traditional image of Apache running on Red Hat Enterprise Linux version 4 is 3GB, whereas the same application in DAB form will be 30MB.

In the event a situation arises requiring other or additional software components, the application still will have on-demand access to them via the repository. However, McLeod notes, this capability is best suited for the testing process, as many users prefer to harden the production-version application by turning off the link back to the repository.

So how does this solve the problems of low utilization and difficult administration? McLeod says the smaller images enabled by FastScale Composer mean less maintenance, fast testing, and utilization rates up to 90 percent. Reduced time to production also is a big draw, as applications and servers can be ready to go in minutes or hours. The longest part of the process, says McLeod, is uploading everything into the repository. A standard OS, he said, takes about 1.5 hours.

As for flexibility, McLeod says FastScale Composer allows for on-demand re-provisioning of servers, and enables any application to run on any machine. He added that Composer works well with third-party load balancers, schedulers and policy engines, and that DABs are small enough to run non-persistently in-memory, which means “all of my processing power is completely available for any application at any time.”

“The idea of dynamic provisioning is to be able to minimize the compute pool of resources that you need in a datacenter and allocate resources on the fly as particular resources are needed, and at the same point, de-allocate resources when they’re no longer needed,” said Chris Wolf, senior analyst with Burton Group. FastScale allows for this functionality, Wolf said, which is a core theme of any vendor talking about the “dynamic datacenter.”

But FastScale’s Composer Suite does not end with physical machines; its Virtual Manager product extends the FastScale approach to virtual environments, as well. Essentially the same thing as Composer, but with VMs, McLeod says Virtual Manager actually allows for users to run significantly more VMs on a single machine than using VMware alone. Tests have shown a 4GB machine that normally could handle only 13 VMs was able to house 42 VMs by leveraging FastScale technology. All of this, says McLeod, can be accomplished without performance degradation, and because the DABs are running in-memory, all 42 can be re-booted faster than bringing up one traditional VM. Users can dynamically deploy and switch between physical and virtual machines, and can build and deploy VMs within seconds.

Another key feature of Virtual Manager, says McLeod, is the ability to build VMDKs. They can be application-specific or uniform, depending on user-defined policies, and are ready for action upon reboot, he added. Because a VMDK is just another VM, it will work with all other VMware tools. McLeod says that Virtual Manager was developed for VMware, but there are plans to make it work with Citrix XenServer and Microsoft Hyper-V, as well.

Aside from being a technology partner, VMware also is FastScale’s biggest customer, said McLeod. However, he added, the company is gaining a lot of traction in the financial services space, where firms constantly are undertaking next-generation datacenter initiatives to increase their competitive edge. One banking customer currently evaluating FastScale has more than 1,300 different applications, and it plans to maximize them by replicating component databases and DABs to datacenters around the world.

Adding his two cents, Burton Group’s Wolf says that FastScale’s Composer Suite currently is ideal for development, testing and training — especially in situations requiring quick repurposing of servers that require bare metal resources. Perhaps indirectly referencing the same financial customers mentioned by McLeod, Wolf added that “some of your more bleeding-edge IT shops that have been running virtualization in production for several years see products like FastScale as their next logical step” in furthering consolidation and automated server provisioning.

Wolf notes, though, that there are — in some cases — drawbacks to FastScale’s “1.0”-level product. For one, such solutions must look at provisioning through the perspective of the entire data path, including networked storage allocation, network access, VLANs, switches, etc. “There’s more to provisioning a server than just putting an image on a server and turning it on,” he said. FastScale will need to improve in these areas eventually, he added, but the product as currently constituted should work alright in situations where networks and storage are configured loosely, and where users are swapping resources that already have been defined. In fact, he noted, this probably will be the typical FastScale deployment.

McLeod, however, is confident that FastScale’s one-of-a-kind approach to solving datacenter woes eventually will win the day in the larger IT world. “We have a very unique and innovative approach to actually managing the software stack, which, really, nobody else deals with,” says McLeod. “They either … have a way of automating the way you do things today, which is moving these big golden images around, [and] people are doing consolidation based on putting virtual machines in, which we think is a great idea, but no one is actually dealing with how you manage, build and deploy your software stacks.”

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