Wellcome Trust Sanger Institute support data intensive research applications with Riverbed
The Wellcome Trust Sanger Institute is one of the world’s leading genome centres. Since its inception in 1993, the Institute has played a key role in the development of genomic science, from sequencing of reference genomes to understanding the role of genomic variation in health and disease. Through its ability to conduct research at scale, it is able to engage in bold and long-term exploratory projects that are designed to influence and empower medical science globally. Institute research findings, generated through its own research programmes and through its leading role in international consortia, are being used to develop new diagnostics and treatments for human disease.
Challenge: Create a resilient website able to cope with hundreds of simultaneous requests
From its campus in Cambridge, UK, the Wellcome Trust Sanger Institute runs 320 configured websites publishing its research findings. When users request data, often running a number of parallel requests to compare with previous reports, the sites experience big spikes in traffic. In a typical month they receive around 80 million hits, with a peak of over 90 million which translates to over 53 million page impressions. “We are carrying out cutting edge medical research which requires a holistic examination of many biological processes across different systems. Our current web infrastructure could only process 400 simultaneous requests. As we service so many domains, an influx of requests could take down the whole campus web infrastructure. Therefore, it was crucial to build resilience into the system to make it more reliable for our users,” explains Dr Paul Bevan, Core Software Services team leader at the Wellcome Trust Sanger Institute.
The Institute runs a bespoke IT system to allow for the dynamic nature of the applications and content. The back-end applications, which are used to process the raw research data, are developed by the biologists and run on blade servers at the in-house data centre. A second data centre provides back-up. Many of the Institutes’s core services – including exchange and Unix mail servers – are virtualised.
“We were using an open-source load balancer to manage the traffic but it had limitations on the number of requests which could be processed at one time, which often saw the back-end applications falling over,” continues Dr Bevan.
“Additionally, to make configuration changes it was necessary to stop and restart the servers which could happen more than once a day; the result was a lot of short-term downtime. We decided to look for an alternative which gave the web team more control, allowing us to be faster and make changes quickly and easily.”
Solution: Institute can now set up sites quickly
The Institute selected Riverbed Technology, supported by network infrastructure and systems integrator IP Performance, which offered the control and flexibility needed with the Riverbed® application delivery controller (ADC).
“Our consultative approach ensures we understand our customers’ unique requirements,” comments Phil Crossley, Managing Director, IP Performance. “By delivering a managed Proof of Concept in partnership with Riverbed, we enabled The Wellcome Trust Sanger Institute to evaluate the unrivalled performance of a Riverbed ADC in both their QA and test environments. As a technology partner, we will always go that extra mile to delight our customers.”
“With the help of our systems integrator IP Performance, we migrated to Riverbed ADC over a period of six months. However, we took our time over the project to resolve any issues before deployment,” explains Dr Bevan. “The approach also minimized downtime, as some services are so critical we cannot have them offline for more than three hours.
“The Riverbed traffic management software was incredibly easy and quick to install – the virtual machines only took ten minutes. Now all of our traffic, both internal and external, runs through Riverbed.”
Benefit : Flexible soluti on that has reduced latency and costs
Two Riverbed ADCs have replaced two network switches and six physical servers. This saves space within the data centre, reduces energy consumption, and on-going costs. The solution then manages 20 database servers and 70 web servers, to reduce latency and ensure constant availability of the Institutes’s sites.
The solution enables the website team to make configuration changes which were previously only possible by reconfiguration of network switches and hence carried out by the network team. Additionally, the traffic management software from Riverbed allows the Institute to set-up sites quickly and because they create so many sites, this is a huge benefit.
In addition, a sophisticated programming language integrated within the Riverbed ADC can be customized and has provided further control to optimize, and route traffic based on the Institutes’s unique requirements.
“Our requirements are unique; firstly because of the dynamic content and volume of requests we face when research is made available, but also because we run a bespoke IT set-up,” concludes Dr Bevan. “The Riverbed traffic management solution works effectively with our applications, providing our web team with the flexibility they need to handle the traffic, scaling up and down as required. It has allowed us to become more agile and this benefit is passed on to the academics accessing our data.”
About IP Performance
IP Performance has been delivering specialised infrastructure solutions and integrating complex network systems since 1994. We deliver Internet/networking services and products as well as professional, certified network management and technical support services - 24 hours a day.
Our clients range from the largest telecommunication service providers, to blue-chip corporates, to the education and local government sectors. We are fully accredited with a small number of carefully chosen partners, who all deliver secure, reliable, high performance solutions.