Phlexglobal gains visibility into Kubernetes clusters

Large-scale clinical studies can take place across multiple countries and involve thousands of patients—for whom massive amounts of data must be closely monitored—it’s easy to see why it can take more than $2 billion and 12 years to launch a new treatment, and why clinical trials account for a third of today’s drug development costs.

Since its inception in 1997, Phlexglobal has been helping life sciences companies streamline this process by enabling them to take charge of their trial master file (TMF)—i.e. the data repository for all documentation related to a clinical trial. Telling the story of how a clinical trial is conducted and managed, a well-maintained TMF is not only the key to proving compliance with industry and government regulations, it also facilitates and improves collaboration among a clinical trial’s many partners.

In the beginning, Phlexglobal was purely a services company: A clinical trial administrator that brought order, stability, and control to organizations’ TMFs. At the time, this was a mostly paper-based undertaking, involving large records offices, bar-coded files, and an ever-growing amount of physical documents. As documents became digitized, paper-based processes evolved into electronic ones, and in 2007, Phlexglobal released its hosted electronic TMF (eTMF) solution, PhlexEview, to support these new processes. Today, that platform is becoming the business—or at least a large and growing part of it.

Phlexglobal Chief Technology Officer Barry Sacks explains, “Today, Phlexglobal is very focused on capitalizing on the opportunity our eTMF platform presents—making it a much more prominent aspect of our business and using it to transform the company from a services business supported by technology to a much larger technology organization supported by services.” Hired in 2017 to help make that vision a reality, Sacks wasted no time moving Phlexglobal IT from an on-premises environment spread across multiple data centers to a hybrid environment anchored by Microsoft Azure cloud. In short order, the company also transitioned to a DevOps model of continuous development and integration, and adopted Kubernetes container technology to automatically manage, scale, and deploy an ever-growing array of microservices.

Proactive enterprise-wide monitoring

Having effected digital transformation and the migration from a monolithic on-premises solution to a hybrid cloud in a short period of time, the one thing still hampering Phlexglobal’s efforts to become an agile organization was the lack of a system-wide view into performance that would enable proactive monitoring.

Liam Corkhill, head of development for Phlexglobal, explains, “At the time we were still very much beholden to the information we could get out of slow query logs and the database, which meant that some of the issues we were trying to resolve had already been reported by a client. Obviously, this was not a good situation, since the client was disgruntled with our application’s performance before we even knew there was an issue.” Luckily for Phlexglobal, Sacks knew just the remedy. “I’d had a lot of success using New Relic monitoring within my own SaaS organization,” he says, “so when I discovered that Phlexglobal didn’t have that capability or toolset, I immediately saw the potential and opportunity that the New Relic platform could provide.”

The monitoring platform proved its worth almost immediately by enabling the Phlexglobal team to understand performance as the company continued its migration from an on-premise environment to its new hybrid cloud. This benefit was quickly passed on to customers. Says Corkhill, “New Relic really allowed us to see into application performance. As a result, we were able to target the slowest-performing aspects of the application, optimize that code, and very quickly see significant improvements in both our clients' visibility into the system and their understanding of its performance.”

While those initial improvements in application wait time justified Phlexglobal’s investment in New Relic, the platform has since paid for itself many times over.

Today, the company is using New Relic to proactively monitor key transactions, baseline performance, and much more. New Relic has become an ingrained and automated part of the company’s software development process—a far cry from Sacks’ early days with Phlexglobal when the company was challenged to do high-volume or stable load testing in an automated fashion.

“We're now using New Relic very heavily as part of our software development process,” says Corkhill. “This means that we can load-test as part of application development. We can use the deployment markers to identify when a code drop has occurred and then look at New Relic APM (application performance monitoring) to see the impact of that code drop—particularly if we're doing a performance-targeted deployment. We can also drill down to see what the worst performing queries are, what's taking the most time, where the slowest transactions are, and so on.”

Kubernetes control

New Relic monitoring has also been a boon to the company as it modernizes its application by adopting a microservices architecture using containerization and Kubernetes orchestration technology. "As we embark on our microservices journey, we can use the correlation tracing capabilities in New Relic to do things like investigate a slow-running transaction on the website by drilling into the microservice and going straight to the trace, or drilling into the database query. Because it’s all joined up, it makes it very easy for us to investigate the entire end-to-end process flow of the application,” says Sacks.

Corkhill is particularly excited about using New Relic’s Kubernetes cluster explorer tool to get a handle on the health of the organization’s Kubernetes environment. “When I first saw it, my initial thought was, ‘This is amazing,’ because it gives us a graphical overview of what’s happening on the cluster at a given time and what the pods are doing. It also enables developers to see how their applications are performing in a containerized manner. This addition means that we don’t have to find other toolsets or learn new platforms to get that information.” 

Ben French, head of infrastructure, concurs: “The visibility into our Kubernetes clusters that New Relic provides has empowered our developers to be a lot more self-sufficient in diagnosing problems—not just in production code, but also in development and test code when they're in the middle of a sprint. That’s hugely valuable from a development standpoint.”

Quick troubleshooting, happier employees

Since deploying the New Relic platform, Sacks and team have watched employee productivity and customer satisfaction soar as application performance has improved and the time required to pinpoint and resolve issues has dropped dramatically. “While our business model is changing, our services organization remains a driving force for the business. Thus, if those employees are sitting idle because they’re waiting for the system to spin and process something, we’re losing money. Thanks to the insights provided by New Relic, we were able to make improvements to the application’s wait time very, very quickly. That in turn led to a huge improvement in the productivity of those employees,” says Sacks.

But the benefits of New Relic extend far beyond Phlexglobal’s services organization. Sacks and team have also seen the IT support function transformed by the improved troubleshooting facilitated by New Relic. “Our clients don’t always provide a lot of detail around issues, but that doesn’t matter so much anymore because with New Relic, our first-line support people can identify exactly where and when a problem occurred, and what the user was doing at the time. That way, when my support team gets the ticket, that information is right in front of them, whereas in the past it would have taken a significant amount of investigation just to obtain the info—let alone solve the problem," says French.

“By allowing us to overlay application performance with infrastructure performance, New Relic has hugely accelerated not only our understanding of the app but also of how the database is performing at an infrastructure level,” says Sacks. “To capitalize on the huge market opportunity ahead of us, we need to be able to run as fast as we can. That’s why we’re always looking for ways to not only optimize our processes but also to ensure that every piece of effort we expend is the right effort expended at the right time to achieve the right goal. New Relic underpins everything we're trying to do as an organization by providing the insights we need to make the right decisions about our business.”