Insight by Leidos

Cloud Exchange 2023: Leidos’ Lakshmi Ashok on monitoring performance of high-level, edge services

As agencies’ hybrid infrastructures make the delivery of high-level and edge services from the cloud possible, IT teams must apply consistent metrics to monit...

Current modes of federal cloud computing sure don’t look like your father’s cloud. From basic data center replacement, agencies have moved to, “Hey, let’s use cloud for all these neat services,” points out Leidos’ Lakshmi Ashok.

Some of those neat services include improved citizen engagement, numerically intensive data analyses (previously done on super computers) and remotely delivered virtual reality training, said Ashok, vice president of enterprise management services at Leidos.

“And finally, we’re talking about edge solutions,” Ashok said during the Federal News Network Cloud Exchange 2023. “From arming warfighters at the edge to act quickly based on information instantaneously or giving a border security agent the right set of tools so that they can figure out whether somebody crossing over is a good actor or a bad actor, based on biometrics that can be processed right there within a handheld system.”

She cited a couple of developments as leading to these capabilities. First up are a plethora of new services from the cloud providers, Ashok said. Second, the move by agencies to compute infrastructure architectures that make use of clouds and of computers on the edge optimal. When taken together, these two developments minimize both application latency and excessive data transfer, she said.

The use of advanced cloud applications might be exciting, but they require careful planning accompanied by the continual use of metrics to measure effectiveness, Ashok cautioned.

She outlined four high-priority metrics that organizations should track:

  • What did the application change for external users? “Building really good customer experience is a metric I would go by,” Ashok said.
  • Were there speed improvements? That means “speed in getting these applications developed and deployed,” she said, adding, “So many policy changes are happening, and how do you make sure that the applications directly reflect the changes in policy as quickly as possible?”
  • Is your security consistent able to halt breaches before they affect service or do harm? Security in the cloud era requires rethinking the perimeter-moat model, Ashok said. “Now, we’ve got to rethink cloud security from a zero trust perspective.”
  • How is the service affecting costs? Cost is something no agency can overlook. “Thinking about cloud as just another data center, and lifting and shifting applications to cloud, is not going to pay off,” she said.

Controlling costs is a matter of prioritizing what must move to or originate in the cloud, and then ensuring application design and hosting patterns minimize data movement or use cloud resources according to demand, not at a never-ending flat rate, Ashok said.

Bringing automation to bear in the cloud

Ashok noted that agencies increasingly combine data analytics with artificial intelligence and machine learning. They apply these techniques to mission-related functions such as fraud detection or making predictions based on trends drawn from data. These strategies also apply to cloud computing itself and offer additional ways to optimize operations, she advised.

For example, “think about looking at workload scheduling,” Ashok said. An organization can configure workloads to start and stop depending on demand, which may vary in a day or throughout the year.

The analysis of cloud logs over time, for instance, might help an agency’s IT team “recognize a pattern of events that can lead to a failure or reduction of service availability,” Ashok said. Then, instead of having an unpleasant middle-of-the-night surprise, the team can anticipate what could go wrong and when and proactively take steps to avoid service hiccups or downtime.

Ashok cited two other data-powered cloud applications: detection of anomalies that forewarn of cybersecurity breaches and external user patterns that point to improvements to code into better workflows.

In such automated environments, agencies will have to move past a lift-and-shift mentality and building new technology stacks in the cloud for purposes of supporting old applications.

“The third way is something I prefer,” Ashok said. “Redefining the business logic and coming up with a great customer experience.” That approach almost invariably requires that organizations “start fresh” and build their applications in the cloud from the get-go, she said.

Leidos maintains a catalog of native cloud services from multiple vendors to ensure that such development in the cloud occurs securely, Ashok said. She said programming in security lets agencies update their processes for change control so that updates can happen continuously and allow for managing changes against the authority to operate process.

“We’re looking at security solutions like ATO acceleration,” Ashok said. “These become critical when we migrate to the cloud.”

To read or watch other sessions on demand, go to our 2023 Cloud Exchange event page.

Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.

Related Stories