Last week SiliconANGLE reported on the Department of Health and Human Services’ rumored action to bring Verizon’s Enterprise Solutions division into the picture in an attempt to improve the performance of the healthcare.gov site. While the value of such a move by the DHHS is still under speculation due to questions surrounding the Verizon/Terremark relationship, one must wonder if any company or “team of experts” will be equipped to solve such a behemoth problem.

Just this week Terremark itself experienced a pair of outages, not necessarily related to healthcare.gov, but it did affect its availability.

“Tonight, Verizon Terremark again experienced network issues in their data center that caused a system outage impacting the federal data services hub and the Healthcare.gov marketplace application,” the official, who asked not to be named, said in an email to Reuters.

Issues clearly continue to pile up, and it seemed as though the infrastructure shouldn’t have been one of those issues in the first place.  Where will it end?

Cash alone won’t fix these problems

$634 million down the drain to a deeply flawed code, on a site that collapses on itself as a result of its inability to handle the amount of data it will be required to manage in order to function properly. A patch effort isn’t going to cut it for this septic code, but beyond a revamped code, are there other options available for healthcare.gov’s repair?

Back to the Future

Earlier this year Americans became aware that the NSA has been collecting huge volumes of data, which included, among other things, call metadata from Verizon customers. In light of the Edward Snowden scandal, the public expressed a deep interest in what such data was being used for, however, little attention was paid to the technology behind the collection of such data. In the wake of healthcare.gov’s failure-to-launch, perhaps we need now steer our focus back to the data collection technology, which may offer a solution to our current Obamacare data overload issues.
The technology behind the NSA’s data mining program came from Accumulo. Accumulo, a scale-out NoSQL database, was developed by the NSA several years ago when the NSA was unable to find a database that met its stringent requirements. Many government-based projects are familiar with such requirements as they often form hurdles when it comes to security and access concerns. Healthcare.gov is certainly not exempt here! On a side note if the NSA was capable of producing a competent homegrown solution, why didn’t we reach out to them in our own time of need rather than offer our souls up to a single-bidder?

Coding, capacity and security are three major issues facing the healthcare.gov site. While we’ve already discussed the need to expunge the website’s septic code, capacity and security are still throbbing issues that remain at large.

A Knight in Shining Armor?

Industry insiders would refer to Hadoop as the de facto framework for Big Data processing. However, due to its popularity, Hadoop has become a prime target for hackers and cyber-criminals. Accumulo, while often run on top of Hadoop, is equipped with cell-level security capabilities that make it much more difficult for targets to penetrate. Accumulo seems to offer a solution that healthcare.gov could benefit from both in its capacity management through Hadoop and with its unique approach to security management.

Recognizing Accumulo’s potential with cell-level security capabilities that can appeal to multiple industries including finance and healthcare, company Sqrrl has pounced on the opportunity. Sqrrl, a startup seeking to commercialize the security capabilities found in Accumulo, has now raised $5.2 million in Series A financing to promote this Big Data security solution. Perhaps Sqrrl is the expert those managing healthcare.gov need to consult?

Machine Data Analytics – Can they help?

It’s been stated that the solution to the website’s issues lie within the system’s own machine log data. According to Sanjay Sarathy, CMO of Sumo Logic, the next-generation machine data analytics company:

“It’s obvious and unfortunate that right now HealthCare.gov is in a reactive mode and unable to course correct potential site problems before they arise and disrupt the experience of the users.”

From Sumo Logic’s perspective, their machine log data contains that exact information on why the site keeps crashing and isn’t functioning at optimum performance. However, they’re not going to know what all that data means without hefty analytics behind it to tell them what’s happening before the end-users experience the issues. “What they need is the ability to proactively uncover relevant data events across their entire infrastructure to help them pinpoint what to fix now and how to avoid these issues in the future.”

Real-time data analysis, cell-based access control and the ability to modify at various points throughout the data management process, these are all features healthcare.gov is crying out for and it would appear these offerings are all available through Accumolo.

Whether Sqrrl can be our King Arthur and Accumolo our Excalibur is another question but all signs of light seem to be pointing in the same direction. If we want a site that can survive the data capacity and traffic required by the new healthcare laws, we need to implement big data solutions and machine data analytics in multiple ways to provide a management solution that will offer security, performance and a long-standing reliable solution for the American people.