Author Simon Lin at the recent EGEE conference in Budapest Hungary. Image courtesy of TWGrid

Around five years ago, we first began to realize that grid middleware and toolkits alone are insufficient to achieve the vision of grid computing. For success, we must focus on deployment, taking advantage of user needs—such as the needs of the high energy physics community and the Large Hadron Collider—to drive a production infrastructure.

On 4 October 2002, the Grid Deployment Board of the Worldwide LHC Computing Grid held their first meeting in Milano, Italy. The meeting was chaired by Mirco Mazzucato and included 26 participants as well as WLCG project leader Les Robertson, computing coordinators from the four LHC experiments, and numerous country representatives.

At this inaugural meeting, this group discussed the goal of providing a common production infrastructure on which to run experimental data challenges: the result was the LCG-1. The collaboration agreed that large-scale production should begin from July 2003, with the aim of achieving real 24-7 production by the end of 2003.

A truly global e-infrastructure will benefit hugely from rapidly growing areas such as the Asia Pacific. Stock image from sxc.hu

Five years on

So, five years has passed. Where are we in the landscape of building such a global e-infrastructure?

In EGEE alone there are over 10,000 users in 150 virtual organizations operating from 240 sites. More resources will accumulate as we approach start-up of the LHC accelerator next year. It does seem we are well on the way to a truly global e-infrastructure.

Driving forward

As one of the earliest Asian participants to join the WLCG production infrastructure, I can also give you the status of progress in Asia and the Pacific.

Asian Pacific partners are beginning to demonstrate a vigorous synergy. Over the last twelve months we have used 3.5 million CPU hours supporting nine virtual organizations. By early 2008, we will have 21 EGEE sites, more than 3,500 CPU cores and close to 2 petabytes of disk space.

Another noted progress is in network infrastructure: the first 10 gigabit network link between Asia and Europe has been operational since July 2007, built by Academia Sinica Grid Computing Center in Taiwan. ASGC is also responsible for the Asia Pacific Regional Operation Center and instrumental in the promotion and dissemination of e-science and grid activities within the WLCG framework.

However, there are still many challenges. One of the major issues for the Asia Pacific is the lack of driving e-science applications that specifically concern countries in that region. To address this issue, in spring 2006 the ASGC initiated the Avian Flu Drug Discovery Data Challenge, coordinating the participation of pioneering Asian Pacific partners alongside many European partners.

Disaster mitigation using grid technology is another area that may better represent the concerns of the Asia Pacific regions, and this and other applications have been incorporated into the EUAsiaGrid project proposal, currently in progress. However, applications that can meet the needs of Asian Pacific scientists are crucial to the long-term sustainability of grid infrastructures in Asian and Pacific regions.

A truly global e-infrastructure will benefit hugely from rapidly growing areas such as the Asia Pacific; such benefits include not only the pragmatic advantage of running the global infrastructure from three major time zones, but also the opportunity to achieve a truly global geographic distribution.

- Simon C. Lin, TWGrid