Network mythbusters
This page answers some of the myths, misunderstandings and questions people have about our network.
Busting myths about our network
Ours is the only network of its kind in New Zealand for research and education. Like other unique things, being different can sometimes pose communications challenges.
These mythbusters address some of the common misconceptions, misunderstandings or questions that inevitably come with such a unique service. Myths, prepare to be busted...
Myth #1: Our network is just a big video conferencing pipe
BUSTED
Our members have been quick to implement high definition video conferencing as it is one of the simplest and most useful applications possible over our network. This is a great first step but it is not the only application by a long-shot...
Our network enables:
- remote control of research equipment in real-time like telescopes, synchrotrons and sensors
- capturing, storage and sharing of huge international datasets for genomics, particle physics and astronomy
- visualisation of historical artefacts
- streaming of high definition multimedia resources
- efficient business administration (data back up, storage)
The network is specifically designed for data intensive, complex research experiments and simulation. The 10Gb/s capacity of the network provides the head-room needed to cater for these high-end uses. Our case studies show you more...
Myth #2: Our network only gets me to my collaborators within New Zealand
BUSTED
Our network extends far beyond our shores. Our international network extends all the way to Sydney and Los Angeles. Our country’s location at the bottom of the world means we need to go to where everyone else is – they will not come to us.
At our landing points at Sydney and Los Angeles, the network connects with over 50 similar research and education networks across the world, giving you access to millions of people and resources without data charges.
Myth #3: We don't need all the capacity the network provides
BUSTED
The network's backbone runs at 10 Gigabits per second – that’s 2500 times faster than standard business broadband. This very impressive capacity is addressing a real need in the education and research community.
Unlike standard commercial networks, we (like every other research and education network worldwide) purposely over-provision capacity to guarantee network reliability and performance. Our netowrk is specifically designed for data intensive, complex research experiments and simulation. The 10Gb/s capacity ensures there is always enough head-room in the network for traffic spikes – even if there are many data-intensive applications running over the network at once.
Over-provisioning is also about future-proofing - ensuring our network continues to be at the forefront of network evolution. Network traffic increased by 130% between June 2009 and September 2010 – from 1 Petabyte to over 2 Petabytes per month. We are now also seeing sustained peaks across individual 10 Gb/s segments of over 2 Gb/s.
Research and education networks around the world are experiencing similar exponential traffic growth - on average traffic volume is doubling every two years. For example, after increasing the capacity of their core backbone from 10Gb/s to 40Gb/s in 2008, JANET the UK’s advanced network is now operating a trial of 100Gb/s in response to continued traffic growth.
Myth #4: We're expensive
BUSTED
In 2006, the Crown invested $43 million in REANNZ to build the network, invest in capability building and support operating costs for the first four years as revenues from the network users (Members) were generated. In 2009, the Crown invested a further $16 million to allow the network to be extended until at least 2013. The Crown committed an ongoing annual contribution in 2011 of $4 million a year.
The network costs approximately $12 million a year to run. This includes the costs of operating REANNZ, the national and international network provider fees, and the ongoing cost of network and service development. There is a $2 million gap annually between our current income and what is needed to sustain the network.
International benchmarking and our own analysis show that we run an effective and efficient research and education network for New Zealand. (In fact, REANNZ costs are in the upper quartile for efficiency when benchmarked across the international community of national research and education networks.)
In the medium term, the key outcome is to make REANNZ financially self-sustaining. REANNZ is currently working collaboratively with core Members, Government shareholders and other key stakeholders to jointly develop an agreed way forward that ensures the network's continuity.
Myth #5: I can buy the same service for cheaper from the marketplace
BUSTED
We provide a type of service at a price-point not available in the commercial market. We cater for a community of research and education super users who have advanced networking requirements not generally satisfied by commercial offerings, while providing plenty of bandwidth for regular users.
It's like a huge private internet. For an annual membership fee, member organisations can use as much bandwidth as they want; implement whatever services they desire; and connect with other members and their international colleagues as often as they wish.
It is possible for individual organisations to purchase high capacity private links from suppliers, but these only work on a one-to-one basis. You would need to purchase a separate link for each organisation you wanted to communicate with. The cost of establishing multiple direct links would very soon become too much for an individual organisation to bear. This is where the benefits of REANNZ’s collective buying power comes into play. By aggregating member demand, REANNZ is able to procure networks and services at a reduced cost – this is particularly important on expensive international capacity.
Myth #6: REANNZ is just another Telco supplier
BUSTED
REANNZ is an expert purchaser of networks and services – arguably the most efficient in the country. As a not-for-profit membership organisation, we work on behalf of our community for the good of our community, aggregating demand to procure the best and most cost effective networks and services from the market to meet their specialist needs. To this end, one of our core business objectives is to ensure that through our actions all of New Zealand benefits from a fairer and more competitive telecommunications supply market.
Myth #7: Connecting schools to the network will interfere with the activities of the universities and CRIs
BUSTED
International studies, reinforced by direct experience from our National Education Network (NEN) trial, show that a nation's schools bring additional traffic volume to a research and education network equivalent to only 1 or 2 large universities.
Schools traffic on the network flows over a separate 1Gb/s virtual LAN (a virtual pipe within the 10Gb/s network core). This means we can easily manage, control and, if necessary, move schools traffic.
Having schools on the network will benefit the universities and CRIs by increasing the ease by which they can communicate and collaborate with each other – streamlining the vertical integration of the sectors within the education and research system. (The New Zealand Vice-Chancellors’ Committee recognises the value of schools’ participation in their 2008 Annual Report.)
From a financial perspective, schools participation will not change the membership cost for other Members. Pricing for schools is currently full costed based on utilisation of <10% of the network’s capacity.
Myth #8: We need QoS (Quality of Service)
BUSTED
There are many different definitions and understanding of what quality of service (QoS) means. On our network, QoS means ensuring there is always sufficient network capacity so that no traffic flow is impeded.
For regular R&E activities, our network's capacity is likely to offer sufficient QoS. However, for some potential very high-volume activities such as the Square Kilometre Array, regular network users would definitely be impacted without some form of QoS intervention.
We will take the approach common in the advanced networking community of seeking to provide separate circuits (including optical circuits) to manage such situations.
We will not implement the 802.1p standards on the core network, but does support the transit of 802.1p header information transparently.
Myth #9: We cannot rely on REANNZ's network because it doesn't have Service Level Agreements (SLAs)
BUSTED
We do in fact have service level targets. However, when our network was purchased, rather than pay in advance for service credits in the case of outages, we saved millions by ensuring the network was built and operated to telco network engineering standards and practices - the same standard they use to support their mission critical corporate and government customers and networks.
This combination of standards and practices, along with the fundamental design of the physical and logical layers of the network, delivers a highly available network service to members. This is reflected in our statistics which show the network’s uptime (availability) over its life has been greater than 99.9% on the national network.
This approach, together with our view that Service Level Agreements can in some cases cause suppliers to focus on the symptoms rather than the root causes of problems, have been why we do not use them on the network.
You can view our national and international network availability figures in our annual Statements of Intent.