This year at AWS re:Invent, New Relic set out to help forward-thinking software teams that want to leverage the Amazon Web Services (AWS) cloud to move faster and innovate more confidently—and become more competitive. The New Relic platform plays a critical role in many of these cloud migration journeys: It helps organizations adopt AWS faster, evolve their software for greater performance by gaining visibility into complex systems, and measure cloud infrastructure and application performance with consistency and precision. All so teams can deliver the amazing digital customer experiences that increasingly determine business success.
This goal was a common thread connecting many of the lightning talks presented this week at New Relic’s two (yes, two!) packed re:Invent booths. These lighting talks drew big crowds by featuring New Relic product experts, as well as representatives from many of our key customers and partners, sharing best practices and requirements for successful cloud migrations. With a special emphasis, of course, on the central role New Relic plays in every phase of a successful cloud migration journey, from initial planning to post-migration.
Instrument your cloud migration in the right place at the right time
Earlier this year, New Relic introduced the Guide to Planning Your Cloud Adoption, a three-phase (Plan, Migrate, and Run) cloud migration framework. Each phase covers a set of steps, key performance indicators (KPIs), and measurement strategies designed to help firms advance through a cloud migration journey. A lightning talk from New Relic Senior Solutions Consultant Boris Gouchev, entitled “Cloud Adoption Best Practices,” highlighted an issue firms often face as they work through this type of framework: failing to instrument in the right place at the right time.
All too often, Boris explained, teams ignore monitoring during the Plan phase and downplay it during the Migrate phase. But teams that apply instrumentation only during the Run phase may be making a serious mistake.
To illustrate his point, Boris recalled one customer that recently migrated to AWS—and learned a hard lesson about the dangers of not monitoring early enough in the process. As the customer’s team prepared its migration effort, Boris said, it discarded its homegrown, legacy monitoring tool; integrated New Relic and AWS; and began moving its workloads to AWS. The team intended to add monitoring only after completing the migration process.
Unfortunately, once the team moved its workloads to AWS, they felt they were getting poor performance from the new cloud services they were using. In fact, the performance wasn’t just disappointing—it was terrible.
Spooked, the team called in an AWS Professional Services architect, who advised them to deploy New Relic immediately on the migrated apps. Ironically, the customer had already purchased New Relic, but the team just hadn’t deployed it yet! After deploying New Relic, it quickly became clear that the performance issues involved the customer’s own code and configurations in their apps—and not the AWS services.
Boris’ story has a happy ending: The team fixed its performance problems and quickly became a happy AWS customer.
The story prompted Boris and his teammates to look into this problem across the entire New Relic customer base. Given the size of a typical AWS deployment (approximately 750 virtual machines), the group concluded that failing to instrument early enough in a migration could add $5 million or more to a typical $31 million enterprise migration project. As little as 48 hours of instrumentation work earlier in the migration process could have spared the customer three months of avoidable, and expensive, rework.
Boris and his team estimated that instrumenting at the right time—during the Planning phase—would:
- Allow firms to decommission data centers up to 25% faster
- Reduce an organization’s average AWS spend by 8% by getting to the Run phase faster
- Reduce by 21% the full-time engineering resources dedicated to a typical migration
Interested in how New Relic and partners can help? Don’t miss the New Relic Cloud Adoption Solution for AWS!
Focus on critical optimizations
The cloud offers great benefits in scalability, availability, and cost savings, said New Relic Strategy Consultant Bharath Raj during his lightning talk, “Host for the Most: Cloud Cost Optimizations.” Maximizing those benefits, however, typically requires post-migration optimizations of both spend and performance.
Once an organization begins its cloud migration, Bharath explained, it’s time to refine KPIs to define usage, user experience, scaling, and application optimizations. Once the cloud migration is complete, the focus can shift to KPIs related to cloud spend and business agility.
Dashboards built in New Relic Insights can help teams visualize these metrics in highly intuitive ways . For example, Bharath explained, to understand the your infrastructure’s workload, it’s useful to identify trends by monitoring the average CPU usage per IP address, by correlating average utilization rate by instance type, or by tracking CPU usage by container as you scale instances.
To track the agility of your business in the cloud, Bharath suggested that teams monitor the highest memory usage of their instances to see where they may need to re-optimize an instance to account for changes in how it’s used. Similarly, it can be useful to track CPU load by instance type or memory usage over time, using these trends to spot anomalies in your cloud infrastructure.
Finally, Bharath said, a good cloud spend and user experience dashboard can capture metrics about your monthly spend, your overall CPU usage over time or by instance type, and your potential for misallocating money on cloud resources.
Supporting a cloud-native evolution
The post-migration, or Run, phase of a cloud initiative is the point where many organizations think they’ve completed a migration journey—but they may still have a long way to travel.
During his lightning talk called “Foundational Elements Supporting Cloud Native Evolution,” Lars Cromley, director of cloud advocacy at 2nd Watch, a New Relic partner and AWS Premier Consulting partner, reminded an overflow crowd that being in the cloud is not a strategic advantage in itself. The value of a cloud strategy, Lars said, depends on what an organization does once it’s in the cloud.
That’s where an organization’s tools, practices, and engineering culture play a critical role in the success of an AWS migration journey. Teams, for example, must learn to work with immutable infrastructure. They must be able, Lars said, to manage applications, services, and software deployments that involve replacing components rather than changing them (a fundamental principle of container-based workflows).
Lars noted that learning how to replace components rather than fixing them, in turn, has an immediate and positive impact on an organization’s mean time to recovery (MTTR). Being “in the cloud” doesn’t create this source of value; it simply sets the stage for a team with the knowledge and tools to take advantage of it.
Second, you can’t overlook what will be a greater reliance on tooling, Lars said, particularly for monitoring your infrastructure. “Visualizing the system and keeping a picture of everything in your head becomes harder,” he added. “The tools that support our applications must have greater capabilities—and those systems, too, must evolve.”
From migration to cloud-native alerting—New Relic gets you there
Regardless of where you are in your AWS cloud migration—starting from the planning phase, looking for post-migration optimizations, or considering the next phase of your cloud-native evolution—these stories are useful reminders that today’s software teams are increasingly reliant on proven best practices. New Relic is an essential component of that process, delivering the observability and insights required to put these best practices to work and achieve measurable success.
Ready for more? Check out all the AWS re:Invent 2018 lightning talks from our New Relic experts.