MORGANTOWN, WV – When a server goes down in a typical office building, the disruption is annoying and frustrating for those affected. But in a healthcare environment, the consequences can literally mean life or death.
That is why IT administrators Farhan Ahmed and Bill White heaved a collective sigh of relief when West Virginia University Hospitals recently installed load balancing technology from Seattle-based F5 Networks in its data center to prevent those disruptions from happening if a server malfunctions or needs maintenance.
“We are on a nearly paperless system, so when it goes down and you don’t have that documentation, a patient can suffer as a result,” said Ahmed, network engineer. “With the F5 box we can take the server down behind the scenes without anyone being affected. We have effectively eliminated any outages the users might see.”
Officially called the BIG-IP Local Traffic Manager, the application delivery networking system is designed to provide sophisticated load balancing and traffic management as well as advanced application security, acceleration, and optimization. The system offers a full set of unified application infrastructure services aimed at giving IT support personnel total control of their servers.
As a Unix administrator, White says the F5 system has made his life a lot easier because now “users are getting in when they need to and if the server goes down, we don’t need an intervention to take care of it.”
The seven-hospital network purchased the Big-IP Local Traffic Manager in September and it was in production the next month, Ahmed said.
“There was a push to have something that would meet future needs,” he said. “With multiple web servers, I realized we needed a load balancer so that if one server went down, traffic would automatically get routed to another.”
Lori MacVittie, senior technical marketing manager for F5 Networks said server disruption challenges are very common and focus mainly on availability and performance.
“The core reasons behind implementing load-balancing technology are to address failure, increase capacity and improve performance,” she said. “If the demand on a service is such that it requires two or more servers to meet with accepted performance, then it is likely that a failure of one of those servers will have a negative impact on availability and performance. Server disruptions can have many causes, including external disruptors, such as attacks. Load balancing mitigates the impact of these disruptions by providing technology that can optimize the resources available and therefore offset the loss of capacity due to disruption.”
Load balancing in implementation is the “virtualization of an application” that can assist in times of disruption by providing a way to seamlessly replace a server with another, MacVittie said. “This gives IT time to address the disruption without impacting end-user availability and performance,” she said. “Finally, load balancing technology when provided by an application delivery system may have additional security features that are able to prevent attacks from reaching servers and causing disruptions.”
Ahmed said he was familiar with load balancing technology from his previous job, so he was sure it would be the answer his team was looking for.
“We really needed to address the potential issues with the server,” he said. “People had to spend more time than usual to get to our application and having to type in their password several times added to their frustration. When we told management about it, they thought it was a good idea and they approved it.”
Several months since it was implemented, the F5 load balancer is properly doing its job – keeping things in balance, Ahmed said.
“Our expectation was that we would need it for the future and so far it has fulfilled the requirement for the present,” he said.
