The Network of Networks

In the Dante headquarters in Cambridge in the UK, Richard Forsyth met up with Dai Davies, General Manager of the GÉANT network soon after it was re-launched in January, this time with an accent on improved service.

At the beginning of 2010, the pan-European GÉANT network – which can transfer huge data quantities between a network research and education institutions – was relaunched with the blessing of the European Commission and a 93 million Euro budget. The same figure was contributed by the Nation Research and Educational Networks (NRENs), who rely on the infrastructure for a variety of R&D and educational initiatives: GÉANT allows large quantities of research to be shared and worked on in real time by different research groups.

“In reality,” says Dai Davies, General Manager of the enormous GÉANT network at Dante (Delivery of Advanced Network Technology in Europe), “this is the network’s eighth version”. The journey of this scientific research-facilitating infrastructure has indeed seen an impressive evolution from the 1980’s and quickly grown into the backbone of major research and development in Europe, for connected institutions. GÉANT is the world’s largest and most advanced multi-gigabit network dedicated to research and education – operating at speeds of up to 10 Gbps.

GÉANT has links to the other major network infrastructures across the world such as EUMEDCONNNECT in the Mediterranean, TEIN in the Asia-Pacific region, CAREN in Central Asia, ORIENT in China, Ubuntunet Alliance in sub-Sahara Africa and BSI in the Black Sea region, meaning millions of researchers can benefit from high speed connectivity in an international context. Also announced by Dante is the creation of the new Cambridge based Network Of Operations – known as NOC, responsible for, and centralising, management of the entire GÉANT network. The NOC will focus on trouble-shooting problems and planning maintenance and upgrades.

“It’s like a heart transplant when you change the network operations centres. This does reflect that network operations is becoming more and more important – it’s not just technology that is important – it’s making that technology work in the way that users want and require and this is the biggest challenge,” says Davies. “If you go back not so long ago to when a network was rationed and expensive, giving anyone connectivity was great, but now connectivity is relatively cheap for quite a lot of Europe and people are finding that even though they have got the capacity available it doesn’t always work well because it is relatively complicated how we connect the systems together and it is certainly not plug-and-play today. Interoperability is still a very big issue at the higher end of performance. By and large the people at the higher end of performance are more demanding but also they are more knowledgeable. But they are knowledgeable about what they want to do and they are not knowledgeable in terms of the way the problems are solved to get there.”

Technical and organisational management are now the focus of Dante, who have become increasingly client focussed: any pan-European system needs to be consistent and requires close monitoring. As Davies explains, the days of giving people technology and allowing them to ‘get on with it’ are over, so a more supportive environment is required. Better troubleshooting and analytical tools are in development, in part to defend against serious issues developing in the first place but also to solve issues with speed and accuracy.

“It becomes increasingly important to monitor the performance in real time the network that you are building for people. You will do it at one level anyway at network operations but this is much more. If something doesn’t work you have to have better analytical tools to understand why it is not working. There are a series of areas in development that we are involved in now, looking at monitoring and looking at security aspects that you associate with networking in general. Once you start to dedicate network resources to people you have to be reasonably sure that the people who are asking have the right to use them. You end up looking much more closely at security issues and authentication issues.”

A network like GÉANT could support some groundbreaking initiatives, notably CERN who will use the network as a means to process the vast amount of data that will be produced in the Large Hadron Collider experiments. However, other sectors are now exploring the effectiveness of using grid technologies to improve communication.

“The most interesting success story has been with radio-astronomy because that was an area where people were carrying out regular communications by magnetic tape,” explains Davies. “Typically astronomers needed to compare the sky in different places from different views. Historically in terms of the cost and the communications and the way it was set up, they would record the data they were seeing with the telescope on to magnetic tapes – on hard discs – and ship them to the central computer to analyse the results. That meant that they were never doing anything in real time because they were collecting data and shipping it off. So by putting that capability on the network it meant that they could do their science in real time and it meant they could do better science and that was a very interesting example from my point of view because it was where the network made a difference and made a fundamental change in the way the science was carried out.”

Grid networking in the biological sciences area is also promising, as it constantly has to share large amounts of data; as is climate research, an area of science that depends on cooperative collection and data analysis.

The value of the GÉANT network becomes more transparent in the context of a serious natural disaster. From the devastating tsunami created near Sumatra in 2004 to the recent Haiti earthquake, the network has been an asset in allowing real-time sharing of satellite imagery and in terms of coordinating information across multitudes of organisations involved in decision-making through disaster management. In some ways GÉANT can be seen as more than simply an enormous infrastructure of fibre optics, and can be perceived as a means of collaborating and resource sharing that can save lives, change the world and rocket our scientific endeavours toward astounding new horizons. The third generation of GÉANT will run through to 2013 when no doubt a new proposal for the next incarnation will be required.

Published: Wednesday, 21st April 2010 by Adelle Kehoe

Comment on this article

Comments

No comments for this article, be the first to comment