Tuesday, November 4, 2014

Overview of 'Open' Software Defined Networking (SDN) for Business Managers


If you don't understand what Software Defined Networking (SDN) is, don't worry about it. When asked, only 10% of 450 information technology (IT) specialists at a recent Network World event raised their hands and said they understood what it was all about.

Obviously, if you’re in IT you will definitely want to know about the topic of Software Defined Networking (SDN). For business managers in most corporations, not so much. However, this article attempts to provide a high level overview of the topic for managers so they don't get lost when the IT staff inevitably starts talking about SDN and why the organization needs to invest in it.

Software Defined Networking (SDN) is a relatively new approach for managing and operation large scale telecommunication networks. SDN involves decoupling control of the network from the physical infrastructure. It allows network administrators in your data center to better manage your telecommunication networking resources that often consist of equipment from multiple vendors.


Background

Sometime in the year 2000, the Gartner Group recognized the emergence of programmable networks as the next big thing for the Internet. There was a growing mismatch between market requirements and network capabilities. In response, the industry started creating the Software Defined Networking (SDN) architecture and associated 'open' networking standards.

The growth of mobile devices, server virtualization, cloud services, and the emergence of the Internet of Things (IoT) are among the trends driving the telecommunication industry to reexamine traditional networking architectures. Many conventional networks are hierarchical, built with tiers of Ethernet switches arranged in a tree structure. This design made sense when client-server computing was dominant, but that architecture is ill-suited to meet the dynamic computing, storage, and communication needs of today's enterprise data centers and telecommunication networks.

'Open' SDN Technology & Solutions

Open source software now plays a permanent role in the world of enterprise IT systems. Gartner forecasts that open source technology will be included in 85% of all commercial software packages by 2015 and 95% of mainstream IT organizations will use a range of open source software components. Currently, one of the fastest growing segments within the world of open source software is Software Defined Networking (SDN). The SDN market is projected to surge from $360M to $3.52B by 2018.

The Open Networking Foundation (OFN) was founded back in 2011 to champion the cause of 'Open Standards' and Software Defined Networking (SDN). In addition to promoting SDN, it also supports the popular OpenFlow specification and communications interface for creating 'open' SDN solutions. Its board members include Microsoft, Yahoo, Facebook, Goldman Sachs, Google, Verizon and Deutsche Telecom.

SDN Benefits

The tangible benefits of transitioning to software defined networking are many. They range from better management control of an organization's telecommunication networks to significant reductions in operational costs over time. Next generation SDN solutions are not only designed to handle the many new technologies and infrastructure requirements associated with the cloud and the Internet of Things (IoT), but they provide a foundation for massive scalability to meet the future needs of organizations well into the 21st century.

Other benefits include more vendor independence, transitioning away from expensive proprietary solutions, as well as increased flexibility and agility needed to innovate and meet future needs that have yet to be determined. In addition increased network performance will also translate into better service to network users - business customers.

According to Amin Vahdat, a Principle Engineer at Google, "The biggest advantage is being able to get better utilization out of our existing lines." Currently, the state-of-the-art in the industry is to run lines at 30% to 40% utilization. With SDN, organizations should be able to run wide area lines at close to 100% utilization.

Conclusion & Recommendations

In traditional data centers, network services have been provided by specific physical devices such as routers, switches, and firewalls. Each of these physical boxes was expensive, complex, physical fixture located in a large data center. All that will change when the network is no longer restricted by the hardware, but lies in the software. The trend is unmistakable. In the future, more and more network infrastructure and services will rely on 'open' standards-based software defined networking (SDN).

If your organization uses large telecommunications networks, better pay attention and plan on funding the your data centers acquisition and transition to the use of software defined networking (SDN) solutions. This applies to organizations in healthcare, education, manufacturing, retail, government and almost every other industry.


Selected Links

OpenDaylight - An open-source platform for network programmability to enable Software Defined Networking (SDN) and create a solid foundation for Network Functions Virtualization (NFV).
OpenFlow - Managed by the Open Networking Foundation, it is an open source communications interface between the control and forwarding layers within an SDN architecture.
Open Platform for NFV Project - A carrier-grade, integrated, open source reference platform intended to accelerate the introduction of new NFV and SDN products and services.
Open Networking Foundaion (ONF) - Dedicated to the promotion and adoption of Software Defined Networking (SDN) through open standards development.
OpenStack - Software designed to control large pools of compute, storage, and networking resources throughout a data center
Project Floodlight - One of the world’s leading open source software-defined networking (SDN) community.
SDN Central - One of the leading centralized source of news and resources for network functions virtualization (NFV) and software defined networking (SDN).

* Also check out PLUMgrid, JedaNetworks, Embrane, Big Switch Networks and Midokura.



Wednesday, October 29, 2014

Ebola in the US: Short on Accountable 'Open' Information, Effective Systems Planning and Decision Making


The CDC issued new guidelines for workers in U.S. hospitals who care for Ebola patients. Even when they follow all recommended guidelines for wearing personal protective gear, they “might not realized they have been exposed.” Statement from CDC, Washington Post October 28, 2014

By Marc Wine

This personal blog relays some of my experiences and efforts related to helping produce a robust and integrated biosurveillance capability for the nation with connections to international disease surveillance systems, in order to provide early warning and ongoing characterization of disease outbreaks like Ebola.

Events in the present Ebola crisis prompt unease that the United States deployment of Web based, standardized population health and biosurveillance information services is fragmented, incomplete and insufficient, prompting me to write this blog.

The United States has made significant progress in public health and medical preparedness since the 9/11 terrorist attacks; yet, poorly interconnected information systems add to our vulnerability to planning and response to viruses like Ebola or enviro virus EV-D68 that threaten the health of large populations.

Today, a gap exists between information technology specialists and public health programmatic or scientific personnel. Overall, their intergovernmental attention to interoperable data management for meeting the challenges of notifiable disease reporting, outbreak detection, emergency response, program evaluation, and public communications in relation to syndromic disease outbreaks, is divided by lack of program collaboration and timely, accurate data sharing.

The will of the nation’s public health decision-makers has been divided between federal, state and local politics that prevent connecting content, analysis and integrated electronic health information systems from every hospital, clinic, county, state and federal departments of health with the people.

Today, the Ebola outbreak that began in West Africa, killing over 5,000 people so far, moved from Texas Health Presbyterian Hospital in Dallas to the Washington, D.C. area with the transfer and now released Nina Pham the first nurse diagnosed with Ebola after caring for now deceased Ebola patient Thomas Eric Duncan, to the situations in New York with Ebola diagnosed Craig Spenser and quarantined Kaci Hickox in New Jersey plus the dozen U.S, military personnel quarantined in Vicenza, Italy following their mission in West Africa.

My sympathies are deep for the people and families stricken by the virus; further, my empathy is heightened for all people who are missing the promise of security and protection from not having a deployed and interoperable nationwide information system for the continuous predictive analysis and communications in the common handling of critical disease and public health threats.

What we do see are the results of the gaps and dangers of the unfulfilled policies of the National Strategy for Biosurveillance (The White House July 2012).

My own observations of the wide-scale implementation of electronic health information systems for surveillance show challenges, barriers and missed opportunities suggesting the U.S. failure to fully harness the laws and comprehensive systems of early warning of health threats and early detection of deadly disease events, including bootstrapping comprehensive and unremitting education especially among health care workers.

Situational awareness in real-time, that ought to have enabled the more appropriate response behavior of the health systems workers in Dallas, Texas for example, remain uncoordinated with federal and global signs of the Ebola crisis. Governments around the globe have stalled, unwilling to recognize this outbreak as the humanitarian, information and resources crisis it is. Even though, we knew about Ebola at least since the 1976 Sudan outbreak, the U.S. is looking months into the future before any efficacious vaccine may be expected.

Looking back, beginning within about three months of the 9/11 attacks, Veterans Health Administration (VHA) Under Secretary of Health Dr. Jonathan Perlin requested that the VA Federal Interagency Health IT Sharing (HITS) program to demonstrate within 90-days the capability for VA to exchange the data of all VA ambulatory lab results system-wide with CDC’s Office of Public Health Informatics in near real-time.

The first goal of the initiative was to electronically exchange the data from VA to CDC plus create algorithms for analyzing the lab test results for flagging any evidence of atypical diseases or syndromic incidences; supporting the monitoring and detection or population disease or biosurveillance outbreaks. The outlier lab tests of unusual and unexplained results would be exchanged back to VA end-users for considering potential bioterrorist or syndromic disease events.

Then working on the mission to plan with the CDC associate director for health informatics, we helped deliver on the Undersecretary’s request for the development of a data exchange and analysis demonstration between VA labs and CDC. Accordingly, It marked the first time VA was to share lab data electronically on a wide-scale with CDC for public health disease monitoring. Although, it was not until about a year later that the VA and CDC could show the initial capability of how to cleanly exchange electronic disease surveillance data, initially via an over-night batch processing solution.

Eventually, this evolved as a part of today’s BioSense 2.0 that provides a mechanism to collect and share information on emergency department visits, hospitalizations, and other health related data from multiple sources, including VA, the Department of Defense (DoD), and civilian hospitals from around the country.

In 2002, I discovered that VA clinical researcher Dr. Sylvain DeLisle of the VA Medical Center Baltimore was awarded CDC’s first grant to mine VA’s VistA electronic health record (EHR) system for the purpose of identifying evidence of influenza-like symptoms that may be indicative of serious population health threats. Dr. Delisle’s results, gained from applying his methodology used to optimize VA’s VistA electronic health record (EHR) data, evaluated the performance of an automated text classifier for syndromic surveillance.

This VA data mining automatically processed VistA health record source documents, patient diagnostics, and was used for informing decisions regarding electronic textual data sources for potential applications with computerized biosurveillance systems. In other words, Dr. DeLisle’s analytic method would automatically flag certain patient diagnostics for potential unexplained disease incidence.

My reaction immediately and vigorously was to help promote Dr. Delisle’s surveillance methodology as an innovative direction toward advancing 'open' solutions, that would be used for generating decision support and knowledge. The analysis of existing patient information could be used to inform the public and health care workers about high potential, critical disease outbreaks across health systems that can be predicted accurately from patients’ electronic health records.

The vision I held then, was that semantically accurate disease predictive analytics can be combined with online learning modules that would be able to inform health workers in real-time how to apply precautions and procedures for emergency response to threatening outbreaks. Now, in the era of Ebola and enviro viruses, it still looks as if the nation’s health system is not there, despite the development of policies and capabilities based on billions of dollars invested over at least a decade-and-a-half.

It appears that CDC Director Thomas Frieden’s answers to questions about Dallas health worker’s lacking a clear understanding about how to handle Ebola is not his part of his responsibility; yet, it represents another view that U.S. leadership with population health and biosurveillance systems is sadly lagging behind policies, plans and political commitments dating at least back to 9/11.

In 2007, my experience working within the Department of Defense (DoD), Telemedicine and Advanced Technology Research Center (TATRC), proved again that the U.S. was planning but not working effectively to produce and deploy an interoperable and integrated data sharing and decision support system for the public’s protection and health care worker’s knowledge of future pandemic or bioterrorist diseases like Ebola.

In the report, “CDC’s Vision for Public Health Surveillance in the 21st Century,” CDC authors emphasized, “…federal, state, and local agencies and health departments have failed to obtain access to desired administrative or survey data…” in relation to collaborating Information from disparate sources or programs that can display patterns of disease that individual program data cannot.

At TATRC the deputy director assigned me to serve as “the belly-button” for promoting innovative health information technology solutions that would be used for managing biosuveillance in educating, informing and analyzing real-time bioterrorist or syndromic disease events. There, I set out to accomplish three goals:
  1. analyze the alignment of the rules and regulations governing how and why population health data are collected and released,
  2. coordinate planning for innovations generating the processes to put the data into a form that can be shared across the Web and between DoD, VA and CDC and
  3. encourage the willingness to use those resources across the public-private sectors.
Then Associate Director for Science, National Center for Public Health Informatics at CDC Dr, Tom Savel visited TATRC, and we agreed to work together on identifying gaps of interoperability between CDC, DoD and VA public health data surveillance systems.

There was no existing useful sharing involving all three of the federal government’s disease surveillance systems across CDC, VA and DoD. My work entailed bringing together managers and leaders of the federal disease surveillance systems from, DoD, VA and CDC for mapping ways to address the policy and technical barriers to cross government data communications about potentially deadly disease outbreaks.

The lesson I learned from facilitating collaborative efforts to meet the needs of public health surveillance programs through collaboration, innovation and open data solutions was that no common standards for communicating, interpreting and educating people from top-to-bottom across the nation were operating; moreover, the efforts just to get the different departments talking to one another proved arduous.

Earlier in 2006, I was invited to attend a small meeting at The White House Executive Office Building where I saw presented the first model for rapid identification of “conditions favorable” for Ebola epidemics using satellite imagery.

As it turned out, this was somewhat related to innovation projects with the TATRC portfolio that I was promoting in 2008. The most forceful realization of this initial disease information satellite project was the possibility of connecting a global biosurveillance system seamlessly to hospitals in America using information technology so that patients would not be seen by American healthcare workers without access to immediate situational awareness of what that patient might have been exposed to while traveling overseas.

With the history, benefits and opportunities that CDC Director Dr. Frieden has behind him, not to overlook the nation’s high priority commitments to systematically inform its health workers and people of the granular details and data about Ebola fully and in advance, how can he not accept responsibility for what happened with Ebola in the Dallas hospital and the incidences involving the subsequent illnesses and gaps overall?

Indeed, an apparent breakdown in the nation’s ability to use electronic health information systems for providing disease surveillance, share open data with interoperability, including situational knowledge in real-time from hospitals across cities and towns, to counties and states, to federal decision-makers and global leaders is contributing to fundamental weaknesses in our global preparedness for future epidemics, which given the interconnectedness of modern life, will likely occur.

Not only does today’s call for answers to the problem of being ready and smart about Ebola cry out for open solutions across all boarders, but my own experience speaking within the Washington, D.C. and global community about the substantial need to be ready with health information systems, shows how little we listen and take action successfully when it comes to national and global governments collaborating effectively.

For example, in 2008 after I was invited to organize and moderate a meeting of top disease surveillance leaders from the U.S. and Canada, including DoD, VA and Canada’s public health departments, hosted at the Canadian Embassy, I became curiously intrigued at the slow pace that America moved with global partners in response to the laws and demands for implementing a reality-based integrated population health disease information network, including virtual reality programs for educating health professionals that CDC itself promoted.

Still, the goal that is evasive in the midst of the Ebola outbreak in the U.S., is that all data potentially relevant to public health surveillance would be harmonized across data systems, interoperable, and easily accessed by the maximum number of users in as timely a manner while protecting confidentiality and privacy of respondents.

More recently, in 2013, I was invited to meet again with CDC officials. This time the initial request was for me to organize a meeting with the CDC public health Community Guide directors and The White House Chief Technology Officer (CTO). The purpose of the talks and was for enhancing the planning and design of a nationwide public health decision support system that would generate usable knowledge, from the federal level with secondary epidemiological and population health research, and deliver it in real-time to all state and local levels of public health directors for their use in planning and preventing serious disease or other public health threats.

On the one hand, the success of my talks with the CDC officials about how to design an infrastructure for an advanced health information decision support system, one that would responsively handle knowledge sharing about an Ebola outbreak for example, in an extreme instance, was noted by their appreciation for learning how to proceed. On the other hand, I am substantially more concerned today than ever before in my experiences that far too few have stepped up in advance to provide the resources and services including technical expertise and political will that are so desperately needed to fight Ebola at every level.

Final Observations & Recommendations

These are selected observations and recommendations for working urgently towards a more responsible network for filling the gaps of systems interoperability, plus sharing health and disease information.

The job of deploying a comprehensive disease surveillance information system is unfinished.

Today, this presents the nation with a critical imbalance from having to react lesser prepared than it could have, as did health workers handling Ebola in Dallas, to benefiting from having an open and uniform infrastructure for surveillance information and knowledge for all.

By identifying, sharing and integrating diverse information sources and expert analysis, col­lectively we will be more likely to identify trends signaling an incident, analyzing prevalence and better able to answer key questions that President Obama and Dr. Frieden have stumbled into.

Recommendations:

  1. Pursue more 'open' biosurveillance activities that purposefully mix and match efforts and the sharing of information between and among Federal, State, local, tribal, territorial, private, nongovernmental, academic, and other national enterprise participants.
  2. Develop connections through collaborative international biosurveillance activities that will accelerate effective response to domestic and international incidents.
  3. Ensure the nationwide, vertical top-to-bottom 'open source' distribution of a cloud-based data integration platform supporting collection and integration of biosurveillance of information from a variety of governmental and other sources, including social media and news reports.
  4. The president and CDC director should be held immediately accountable for comprehensive disease security and data sharing through the elimination of differences in coding, formatting, definitions, and methods that differ substantially or in the ways data are stored in incompatible formats.
  5. Directors of the states departments of public health should be held accountable for limiting sharing disease surveillance data by the lack of user-friendly data dissemination tools or adequate and detailed documentation and distribution.
  6. The White House and Congress should lead a review of all funding steams and mechanisms the affect how disease surveillance data are collected. All data, analytics and systems deemed to be essential for public health safety, security and learning should be considered open source for the purposes of national security and disease protection.
  7. Data-use agreements should be shared widely to provide models for others interested in sharing data; data sharing should be promoted by developing supportive funding mechanisms, devoting resources, fostering partnerships and centralizing support; and methods and procedures should require open standardized across datasets.
  8. The White House and CDC should immediately re-commit their agencies and departments to assessing the utility of having surveillance data directly flow into information systems that support public health interventions and information elements or standards that facilitate this linkage of surveillance to action and improving access to and use of information produced by a surveillance system for workers in the field and health-care providers.
  9. Public health and hospital electronic health records (EHR) systems should accelerate their capacity to process both structured and unstructured data access and standardized interpretations faster than HHS ONC Meaningful Use Stage 2 and 3 criteria would require.
  10. Concerns related to disease managements best practices, data quality, data standardization, process automation, work flow design, and system validation all need to be addressed. The need to use new and legacy systems in parallel for a period must be considered and planned for, including the challenging process of transitioning users off legacy systems. Therefore, emergency federal, state and local resources should be made available immediately and effectively in order to establish accountability among all the nation’s hospital networks for identifying and handling Ebola plus other population-level disease threats.

Finally, the future Learning Health System Governance and Policy Framework that is being planned with leadership from the University of Michigan Department of Health Informatics should include the real-time, anywhere, anytime decision-support that the people around the globe will demand in relation to public health events.

References for More Information:



Marc Wine is a senior health systems and health information technology adviser in Washington, D.C. and co-author of the benchmark book, Medical Informatics 20/20: Quality and Electronic Health Records through Collaboration, Open Solutions and Innovation, ( Jones and Bartlett 2007)


Friday, September 12, 2014

Understanding 'Open' Terminology

Having heard so many people using the terms “open systems”, “open computing”, and “open source” interchangeably, believing they all mean the same thing, it seemed appropriate to  write a short blog defining some of these terms and soliciting input on other ‘open’ terminology.

In general, the term “Open” often refers to initiatives whose inner workings are exposed to the public and are capable of being further modified or improved by any qualified individual or organization. “Open” is the opposite of “proprietary” or “closed” environments. In the case of software, this would mean that the “source code” is either open for all to access such as the Linux operating system or closed systems such as Windows  where only Microsoft programmers are able to change the source code. 

Other ‘open’ terminology often loosely bandied about include:
  • Open Source Software (OSS) - OSS refers to a software program in which the source code is available to anyone for use. It can be modified by anyone from its original design free of up-front license fees. The source code is available for review, modification, and sharing by the at-large community.
  • Open Standards - The set of specifications developed to define interoperability between diverse systems. The standards are owned and maintained by a vendor-neutral organization rather than by a specific commercial developer.
  • Open Systems - Hardware and/or software systems that use or adhere to open architecture and standards that support interoperable to some degree. See http://en.wikipedia.org/wiki/Open_systems
  • Open Architecture - An Information Technology (IT) architecture whose specifications are open and available to the public and that provide a platform that enables continued evolution and interoperability. See http://en.wikipedia.org/wiki/Open_architecture
  • Open Access - Providing free and unrestricted access to journal articles, research findings, books, and other literature. See http://www.soros.org/openaccess
  • Open Data – Data that anyone is free to use, reuse and redistribute without restriction. For more detail, see http://opendefinition.org.
  • Open Data Format - A standard way for describing data formats, per the “Open Data Format Initiative (ODFI)”, and a program to validate that a data file is “ODFI compliant”. See http://en.wikipedia.org/wiki/OpenDocument
  • Open Community - An environment in which the creative energy of large numbers of people is loosely coordinated into large, meaningful collaborative projects and generally avoids the traditional closed organization structure many are used to seeing in the private sector.
  • Open Computing - This is a general term used to describe an “open” philosophy in building information technology (IT) systems. It represents the principle that includes architecture and technology procurement policies and practices that align IT with the goals of an open interoperable computer systems environment.
  • Open Knowledge - An open system of knowledge transfer using the Internet and other information technologies to share best practices, emerging practices, knowledge and innovations within one or more “Community of Practice (CoP)” or across organizational boundaries. Visit http://okfn.org
  • Open Publication License (OPL) - This is a license used for creating free and open publications created by the Open Content Project. Other alternatives include the Creative Commons licenses, the GNU Free Documentation License and the Free Art License. See http://opencontent.org/openpub/
  • Open Source Hardware - Hardware whose design is made publicly available so that anyone can study, modify, distribute, make, and sell the hardware based on that ‘open’ design. See http://freedomdefined.org/OSHW
We are now seeing the emergence of new, related terms like ‘Open Culture’ and ‘Open Society’ as more people and organizations around the world adopt ‘open’ technologies and solutions and embrace the philosophy behind them.

Have you heard some other ‘open’ terminology being used that you can take a shot at defining and share with us?