When Time Is of the Egress: Optimizing Your Transfers

By Andrew Kaczorek and Dan Harris

July 31, 2012

Traditionally running scientific workloads in AWS provides a diverse toolkit that allows researchers to easily sling data around different time zones, regions, or even globally once the data is inside of the infrastructure sandbox. However, getting data in and out of AWS has historically been more of a challenge. The available resources are still evolving and those pesky laws of physics tend to get in the way. Considering the rise of enterprises utilizing cloud for larger data and compute needs and the complexities that come with it, we thought it would be helpful to offer tips on optimizing ingress and egress transfers.

Within scientific computing there is a massive disconnect from theoretical conversations and the real world of data movement. We recently performed a data transfer to Amazon’s Elastic Compute Cloud using their Import/Export service. The service allows customers to mail in data on physical media which is then placed into a S3 bucket or EBS volume of their choice. As an experiment to compare this transfer to network-based transfer mechanisms like multi-stream upload to S3, we recorded all the time it took to prepare and ship the drive to Amazon.

There were several steps to transfer the 317 GBs of DNA sequence data into EC2:

  1. Installed AWS Import/Export command line tools.

  2. Created an Import job using AWS command line tools including a manifest and signature.

  3. Realized that the drive is an ext3 file system (and mounting ext3 on OS X is non-trivial).

  4. Created an Ubuntu virtual machine.

  5. Mounted the drive on the Ubuntu VM and wrote the signature file and manifest to the drive.

  6. Physically labeled the drive with a transfer ID that was provided by the registration process.

  7. Packaged and addressed the drive with a specific address that was to be used for the shipment.

  8. Headed to the local FedEx and shipped the drive overnight.

  9. Waited….

  10. Viewed completed transfer logs.

The next step had us moving the data from S3 to an EC2 instance to use it in a computation run. Direct to EBS snapshot is an option, but due to its higher costs as an image of the drive, the unknowns associated with the newness of the feature, and the constrains to the specific content of the file system, we decided against it.

Table of Shipping and Transport Times:

Prepare Drive

3 hr (concurrent with other project work)

Drive Shipped

4:12 PM EST (FedEx log)

Drive Arrives IAD

3:20 AM EST (FedEx log)

Drive Arrives at Amazon facility

9:45 AM EST (FedEx log)

Drive accepted by Amazon

1:13 PM EST (I/E toolkit log)

Data transfer begins

5:40 PM EST (I/E toolkit log)

Data transfer completes

9:17 PM EST (I/E toolkit log)

Here is a summary of the entire activity:

Total time to transfer 317GB

32 hours

Extrapolated total time to transfer 1TB

39.8 hours

Throughput of active AWS transfer

199 Mbps

Active AWS transfer of 317GB

3.6 hours

Extrapolated active AWS transfer of 1TB

11.4 hours

Overall throughput of 317GB transfer

22.5 Mbps

Extrapolated overall throughput of 1TB transfer

57.2 Mbps

This import job was compared to the results on some recent multi-stream upload tests performed with an envy-inducing 5 Mbps upload speed compared to 1 Mbps.

File Size

Transfer Time

Avg Speed

250 MB – one thread

413 seconds

.605 MB/sec (4.84 Mbit/sec)

250 MB – 30 threads

412 seconds

.606 MB/sec (4.84 Mbit/sec)

1 GB – one thread

1,695 seconds

.604 MB/Sec (4.83 Mbit/sec)

1 GB – 30 threads

1,693 seconds

.605 MB/sec (4.84 Mbit/sec)

We were able to saturate upload bandwidth and ingress at customer sites, which have much higher outbound data rates in the 50 Mbps range. Further, if there’s a bottleneck for delivering data over the wire it’s on the source end and not on the EC2 end of the line.

The results showed that 50 Mbps of upload speed could saturate a company’s network therefore throttling transfer at 70 percent total bandwidth for an outbound rate 35 Mbps. Interestingly, the transfer speed is faster than the Import/Export service. This shows that almost 500 GB could be moved in the same time it took to transfer by shipping the drive. This drive wasn’t filled to capacity and the theoretical Import/Export throughput would use a full drive by extrapolating the time to load 1TB. Loading that extra data would take about 8 more hours and increase the throughput of the Import/Export approach to 58 Mbps. The rate could also increase if the time it takes to prep the drive was reduced.

What we found from our experiment is that the nature of your workflow should be considered when deciding which transfer method to use. If producing a constant flow of data at a rate that matches the allotted upload bandwidth, streaming over the network is a better option. On the other hand, if there is a large, pre-existing data set and no time to wait for it to upload consider using Amazon’s Import/Export service.

Initiating a transfer entirely in software and having the data eventually make its way into the cloud without getting up from your desk is not always practical. For example, a 317 GB payload would take approximately 30 hours to transfer to AWS if using the Import/Export job approach and 30 days to import 1 Mbps uplink was saturated 24/7. Given a typical enterprise uplink of 50 Mbps, the tables would be turned. Let’s not forget non-technical factors involved in the use of the Import/Export approach such as the hassle handling USB drives, cardboard, packing tape, and cranky shipping depot employees.

Lastly, if the over-the-wire transfer is projected to take longer than a business week, use an AWS Import/Export job instead. AWS Import/Export is an extremely viable way of managing the ingress and egress of data until bandwidth becomes more ubiquitous and plentiful.

Editor’s Note: The original byline was incorrectly attributed to Cycle Computing CEO Jason Stowe.

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