GCJ: What, in your opinion, was the most influential grid story (or stories) in 2006?

Crumb: I can't point to a single most influential story but rather a series of grid wins in research and enterprise which leads me to believe that grids are crossing the chasm from HPC and related line of business applications (Finance, Pharma, etc.) and moving mainstream. In the process they have ridden the "hype curve" from wild exuberance to disillusionment and are now building back credibility one successful deployment at a time. This is true not only in North America, but in Asia - where over 2500 participants show up for a Japanese GridWorld event - and in Europe - where the European Commission continues to pour millions of Euros into grid deployments.

GCJ: What do you predict will be the leading story or stories for Grid in 2007?

Crumb: I think the leading story for Grid in 2007 will be the industry sorting through how the concepts of virtualization, automation, service orientation and grid influence enterprise IT architecture and next generation data centers. During 2006, the industry was "storming" on this issue. In 2007, the industry will be "forming" and possibly "norming" on how these technologies fit together and compliment each other on our overall journey toward a next generation of distributed computing. I think the industry will come to the realization that all of these technologies are complementary, not competitive - and simply represent different "parts of the elephant", as Ian Foster suggested in a recent paper on service orientation.

GCJ: On to OGF, we heard from Mark in August, what's happened since then?

Crumb: Since Mark's interview in August, we have: (a) completed the final phases of the GGF/EGA merger integration and launched OGF at GridWorld 2006; (b) continued to progress the standards and best practice work of OGF, (c) planned and begun executing our OGF19 and OGF20 events, and (d) launched several interesting projects that will be delivered during the first part of 2007.

(a) We completed the final phases of the merger integration - a set of activities that my entire Operations team was happy to have behind them. Launching the new organization during GGF18 at GridWorld provided a great opportunity to engage with our community in celebration of past successes and our future opportunity together. Over 40 organizations representing research and enterprise showed their support for OGF's mission by becoming members, which is very encouraging.

(b) With regards to standards and best practices, GGF/OGF produced more documents (over 25) during 2006 than at any time in our history. This includes recently publishing 3 very significant Data Area recommendation documents. And of course we just returned recently from SC06 where we held a very successful interoperability demonstration of our OGSA HPC Profile which enables interoperability between various resource managers based on a combination of web services standards and OGF specifications (JSDL, BES).

(c) We have spent quite a bit of effort planning OGF19 and OGF20 and strategizing on improving our events in general. OGF19 is being held in Chapel Hill, NC, USA, January 29 - February 2, 2007. This promises to be a very interesting event for the Globus Community. We have two day-long workshops planned - Web 2.0 and Grid and Federated Identity and Grids. We are also experimenting with providing a "software providers forum" for Globus users (e.g. TeraGrid, EGEE, etc) to interact with the Globus team (and other providers) through a series of dedicated sessions within the OGF schedule. Software providers such as Globus are critical partners of OGF and our community and we hope to make "software provider forums" an ongoing fixture of OGF events throughout 2007.

We are also hard at work on OGF20, being held in Manchester, UK, on May 7-11, 2007. OGF20 is collocated with EGEE's 2nd User Forum and hosted by UK e-Science and the University of Manchester. It will include an extensive enterprise grid track in addition to our usual group sessions, workshops and forum activities.

(d) Finally, we are completing our first public draft of the OGF Technical Strategy and Roadmap document. This document is being championed by our newly established Technical Strategy Committee and should be available for public comment during mid January just in time for some interesting discussions planned to take place during OGF19. We anticipate that this initial version will spark quite a bit of discussion and input and we are hoping that many within the Globus community will provide input and engage in a lively set of discussions. .

GCJ: Any growing pains or pleasant surprises since the merger?

Crumb: I think my biggest surprise was the commitment shown by our participants as evidenced by their incredible productivity during a significant organizational transition. This goes to show that our "open forum" is still very much supported by the grass roots commitment of people wanting to see grid become a "mainstream" technology. There have been some growing pains along the way, but our core leadership is adapting and will continue to lead us toward a future where we continue to produce relevant, high-quality specifications and best practices for grids.

GCJ: Tell us about your big announcement at SC06 and what happens next with OGF Specifications.

Crumb: The HPC Profile demo highlighted the value of OGF standards in deploying multi-site, multi-vendor grids for HPC customers and illustrates the progress being made by OGF and our members to produce specifications that enable interoperability and accelerate adoption. The OGSA HPC Profile began as a set of observations from several leaders within the OGF community that successful standardization efforts often address small, well-defined and modular subject areas while taking into account broader and sometimes longer term goals. By starting with simpler designs that can be enhanced over time, fast progress can be made and capabilities can then evolve in an incremental fashion. Bringing together this fast-track design approach with existing specifications already broadly adopted within the industry (WS-I Basic Profile) and ones that were maturing rapidly (OGF JSDL and BES specifications) enabled the OGSA HPC Profile to gain traction quickly. From the germ of an idea, to the establishment of the OGSA HPC Working Group, to the demonstration at SC06, it all came together very quickly (about 5 months). Looking ahead, the groups managing the specifications involved in the demo are taking lessons learned, refining the documents, and submitting them for comment and publication. More broadly, OGF is looking to repeat this "fast path" process in other relevant areas.

GCJ: Tell us about some of the other OGF grid projects that have made significant headway this year.

Crumb: OGF has continued to make good headway throughout the year even as we completed the merger process. As I indicated above, we published over 25 documents and have seen many of our specifications gain interest and traction within the grid community (e.g. DRMAA, WS-DAI, JSDL, BES, CDDLM, OGSA HPC Profile, etc. ). Our GIN project (Grid Interoperability Now) has also been gaining momentum throughout 2006. The purpose of this group is to organize and manage a set of interoperation efforts among production grid projects by focusing on a small list of essential services needed for interoperation across these production grids. Obtaining workable solutions in specific areas such as job description and execution, data location and movement, authentication and authorization, is critical OGF's future standardization priorities.

Another project that has begun to take shape within OGF is our Technical Strategy and Roadmap project. A key change resulting from the merger was the creation of a Technical Strategy Committee (TSC) within OGF to document our directions through the publishing of a Technical Strategy and Roadmap (TS&R). The first version of our TS&R document will be published in January 2007. At GGF18 during GridWorld, we began having a series of outreach and alignment sessions with key stakeholders within the community to gather requirements. Moving forward, we will be holding these "requirement summits" at each OGF event and looking for opportunities to utilize the output from these sessions to shape and document priorities, gaps and opportunities through the roadmap process and the TSC.

GCJ: What about case studies? Can you give a specific example of a business using grid and what benefits they realized?

Crumb: We had plenty of examples presented at our recent GGF18 event at GridWorld. These included case studies from a number of different commercial settings including risk analysis, engineering design, simulation, drug discovery, and the like. Frankly, these "early adopter" stories are sort of old news and I find more interesting newer case studies where grids are helping organizations with multiple data centers establish access to distributed data. For example, there are case studies where customer service departments (a far cry from the standard engineering application of grids) are significantly enhancing their core processes because of what grids have enabled them to do.

GCJ: What about next year, what's on the docket, what should we be looking for?

Crumb: Too much to detail them all here! However, in early 2007 you will see the initial publishing of the OGF Technical Strategy and Roadmap. We will be experimenting with our Software Forums that I mentioned earlier in this article. You will also be seeing interoperability demonstrations and fests as we continue to mature specifications.

close window