January 25, 2010
Cloud computing is a new computing paradigm for many but for the rest of us it is simply today's version of timesharing -- Timesharing 2.0. On-demand or pay-for-usage has been the norm for many HPC organizations for several decades. These users either never had the budget for their own computing resources or the project only needed limited access to powerful compute resources.
In that sense, HPC users, like the biggest commercial users, already trust cloud computing with their proprietary applications, data and results. They have been using pay-for-usage services for years and many have evolved from the early days of timesharing. In many cases, supercomputing centers and government research labs provide the compute resources. If you are a commercial HPC user in oil & gas, financial services, manufacturing or other industry, the compute resources will probably be found in the corporate datacenter.
This HPC user community has pioneered the necessary tools to allocate, measure and control access to specific users and projects while protecting the users from unauthorized access or modification of applications and data or malicious erasure or premature disclosure of results. This community also developed sophisticated accounting and charge back software that kept track of everything from CPU cycles to memory usage to access time and storage used. Suffice it to say that HPC users are well ahead of their counterparts in the commercial datacenter, and the latter would do well to look toward the former for some guidance in this area.
Without a doubt, the biggest challenge to cloud computing is security -- the dark side of the cloud. In the cloud paradigm, the user community does not or should not care about the physical side of business operations. In most cases, the physical infrastructure is housed, managed and owned by a third party, and you pay for resources used just like the electric and gas utilities. Despite all these wonderful capabilities and features, security remains as much of a concern for the HPC community as it does for the consumers concerned about protecting their identity and credit card information.
Imagine for a moment the business ramifications if the results of critical drug data or aircraft design were changed and compromised by malicious activity or they were released to the world prematurely. The real and intangible costs to your company can be devastating.
Threats to the network and information security have been occurring for decades, nothing new. However, the complexity and scale of attacks are rising at an alarming rate, presenting organizations with a huge challenge as they struggle to defend against this ever-present threat. Today, cybercrime is more lucrative and less risky than other forms of criminal activity. Threat levels and attacks are on the rise, striking more and more businesses. Estimates for disruption, data theft and other nefarious activities were pegged at a staggering $1 trillion for 2008. Certainly more than a round-off error!
Just this month, the news headlines in CNET News include "Google China insiders may have helped with attack" and in the Wall Street Journal: "Fallout From Cyber Attack Spreads." CCTV.com reported, "China's largest search engine paralyzed in cyber-attack…." And a ZDNet headline on Jan. 21 read: "Microsoft knew of IE zero-day flaw since last September."
In July 2009, the associatedcontent.com headlines read "Near-Simultaneous Cyber Attacks Down U.S. Government Websites." The article reported that the attack targeted the "White House, Pentagon, NYSE, Secret Service, NSA, Homeland Security, State, Nasdaq, Treasury, FAA, FTC, and DOT Websites."
The low risk and low-cost of entry of cyber-crime make it an attractive and lucrative "business." Cloud-based computing exacerbates the situation by facilitating access to increasing amounts of information. IT organizations have a hard enough time defending their in-house private cloud resources. Companies offering public cloud, pay-for-usage models are faced with a more difficult challenge since they must serve multiple organizations on the same platform. At the same time, there is an opportunity for innovation of flexible cloud-based security service offerings.
The criminal element employs powerful tools such as botnets, enabling attackers to infiltrate large numbers of machines. The "2009 Emerging Cyber Threats Report from Georgia Tech Information Security Center (GTISC)" estimates that botnet-affected machines may comprise 15 percent of online computers. Another report compiled by Panda Labs estimates that in the second quarter of 2008 10 million botnet computers were used to distribute spam and malware across the Internet each day. With the growth of the cloud paradigm, more and more mission critical information will flow over the Web to publicly-hosted cloud services. The conventional wisdom of defending the perimeter is insufficient for this dynamic distributed environment. One element in common across commercial enterprise applications is that users must consider security before signing up for public cloud services.
During SC09, I met with many of the HPC infrastructure vendors and also spoke with some real-world HPC cloud users about the concerns they have using cloud computing for their workloads. (This was not a structured industry survey.) Some did express concerns about security but mainly in the context of using public cloud resources versus their private cloud resources. However, they also expressed concerns about transitioning their HPC workloads from in-house resources to external public cloud resources, as it is a very different scenario and from commercial workloads. From a security standpoint the concerns ranged from unauthorized access to exposure of critical information to malicious activity. Additional concerns include the movement and encryption of data to public clouds and the subsequent persistence once workloads have been completed. Has the data really been deleted? It is all about the data integrity.
HPC users often have many options available for running their workloads. For example, an academic user may have access to in-house central computing resources shared between multiple departments, or even access to large-scale supercomputing centers. In this environment the user data, results and applications are still very much 'in-house' and even though there is some security risk, the users are better protected in this environment. HPC users in private industry, especially those in large-scale multinational companies, may have the option of private clouds available for their workloads, and like HPC academic users, have fewer security concerns. However, if the HPC user is looking at commercial third-party cloud providers of public clouds, whether it is Amazon's EC2, Google's App Engine or better still, HPC-specific cloud vendors, they should spend the time to ensure that these vendors fully address their security issues, encryption, and persistence.
To those organizations that do not have internal private clouds and want to use cloud computing from a third party vendor, I recommend you consider the following five security evaluation criteria:
Cloud computing is not so much a new technology as it is a new delivery model, but its impact will be enormous. Research firm IDC estimated that worldwide cloud services in 2009 were $17.4 billion, and are forecasted to grow to $44.2 billion in 2013. The economies of scale and centralized resources create new security challenges to an already stressed IT infrastructure.
This concentration of resources and data will be a tempting target for cyber criminals. Consequently, cloud-based security must be more robust. Spend the time to evaluate the security and make sure it is designed in and not added on after a breach. Partner with a trusted vendor. And if in doubt, seek advice.
About the Author
Steve Campbell, an HPC Industry Consultant and HPC/Cloud Evangelist, has held senior VP positions in product management and product marketing for HPC and Enterprise vendors. Campbell has served in the vice president of marketing capacity for Hitachi, Sun Microsystems, FPS Computing and has also had lead marketing roles in Convex Computer Corporation and Scientific Computer Systems. Campbell has also served on the boards of and as interim CEO/CMO of several early-stage technology companies.
May 23, 2013 |
The study of climate change is one of those scientific problems where it is almost essential to model the entire Earth to attain accurate results and make worthwhile predictions. In an attempt to make climate science more accessible to smaller research facilities, NASA introduced what they call ‘Climate in a Box,’ a system they note acts as a desktop supercomputer.
Read more...
May 22, 2013 |
At some point in the not-too-distant future, building powerful, miniature computing systems will be considered a hobby for high schoolers, just as robotics or even Lego-building are today. That could be made possible through recent advancements made with the Raspberry Pi computers.
Read more...
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
Read more...
May 15, 2013 |
Supercomputers at the Department of Energy’s National Energy Research Scientific Computing Center (NERSC) have worked on important computational problems such as collapse of the atomic state, the optimization of chemical catalysts, and now modeling popping bubbles.
Read more...
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
In this demonstration of SGI DMF ZeroWatt disk solution, Dr. Eng Lim Goh, SGI CTO, discusses a function of SGI DMF software to reduce costs and power consumption in an exascale (Big Data) storage datacenter.
The Cray CS300-AC cluster supercomputer offers energy efficient, air-cooled design based on modular, industry-standard platforms featuring the latest processor and network technologies and a wide range of datacenter cooling requirements.