How Service Mesh Can Reduce the Cost of Running Modern Applications
Flynn, a technical evangelist at Buoyant breaks down why service mesh technology provides features that are critical to your application and how it can reduce the cost of running modern applications.
If you run microservices, you’re familiar with the advantages they bring to your release scheduling and the disadvantages of having everything involve the network. Where monoliths can use secure, fast function calls, microservices communicate via the inherently unreliable and insecure network. We need new capabilities to make this communication layer visible and manageable.
Those new capabilities are embodied by the service mesh: a dedicated infrastructure layer devoted to the reliability, security, and observability of service-to-service communication. A service mesh enables engineering teams to manage communication at a platform level, without needing to drag developers into every issue and has become a critical infrastructure component in any Kubernetes installation.
Service Mesh Adoption is Skyrocketing
In the last two years, service mesh adoption has skyrocketed. According to the Cloud Native Computing Foundation (CNCF) 2022 service mesh report, some 70% of organizations within the cloud native community run a service mesh in production or development, with another 19% evaluating a service mesh. That number has nearly doubled in the two years since the 2020 report, which found only 42% running or evaluating a mesh.
This rapid growth is a reflection of the value that early adopters find when they start using a service mesh. Costs can escalate quickly in digital transformations and IT modernizations, especially when they involve complex, multi-cluster Kubernetes installations across hybrid and multi-cloud environments. Service meshes allow engineering teams to reduce costs by taking advantage of the mesh’s observability, reliability, and security capabilities.
Cost Reduction with a Service Mesh
The first important thing to realize is that service meshes provide critical features to your application. Observability, reliability, and security are not optional: any production application must have these features, so if your infrastructure doesn’t provide them, they must be built and maintained by your developers instead. It is dramatically more cost-effective to let your developers focus on business needs rather than on things the infrastructure can provide.
Next, realize that nothing about Kubernetes at scale is straightforward, and service meshes can bring their complexities to the table. That creates a need for high-quality, marketing-BS-free educational programs. Taking advantage of those will speed up adoption, ultimately saving unnecessary trial and error costs.
But even open source has a cost. Implementing and managing it does require engineering hours. A managed service mesh might be a great option for those who want to save even further or don’t have the bandwidth to operate it.
Developer Productivity
Properly handled, modern cloud-native tools like service meshes can radically improve developer productivity while reducing the workload placed on the developers. For example, if you’re not using a service mesh to provide observability, your developers must add custom health checks, metrics endpoints, and debugging hooks into each and every microservice, by hand.
This approach is slow, expensive, and error-prone. Getting every service covered by identical metrics and hooks is also impossible. Letting a service mesh handle observability instead means instantly getting uniform coverage of your services without asking your developers to spend any time. Likewise, a mesh can effectively manage encryption, retries, and the four golden metrics..
Why put that burden on your developers if a service mesh can do all that uniformly across all services from a platform level? Without worrying about all these features, your developers have more time to focus on generating real customer value by solving real business problems.
Service meshes also allow developers to reliably and automatically diagnose service traffic problems consistently across applications, again freeing up developers to spend more time working on solving important business problems. Microsoft’s XBox Cloud team, for example, saw a saving of $40,000 per month from switching from their observability tool to the Linkerd service mesh. This highlights how large organizations with complex infrastructures can infinitely scale their services while reducing costs and giving engineers significant hours to work on tasks that generate value for the business — a significant operational win.
Secure Network Communication
Service meshes can also help reduce costs even further when it comes to security by transforming each Kubernetes cluster into a secure network. A service mesh encrypts traffic via Mutual Transport Layer Security (mTLS) and upgrades all connections to the latest TLS standard, protecting communication while potentially improving normal HTTP communications performance. Implementing these critical capabilities can be done with no additional configuration and no operator overhead.
While this is another way in which a service mesh frees developers to focus on business problems regarding security, there is an additional benefit. It can be hard to get security right: even small mistakes can leave your services vulnerable to subtle attacks. Moving this critical aspect of microservices into the infrastructure can provide you with much better protection than relying on every developer to get it right.
See More: Intrusion Detection System vs. Intrusion Prevention System: Key Differences and Similarities
Zero-config + Productivity – Cost = a Win-Win
For those working with microservices architectures and Kubernetes, service meshes are a boon for efficiency, developer productivity, and reducing operational costs, all while promoting reliability and security. Engineering teams must allocate more time to coding without a service mesh, application build times take longer, and traffic network management is more complex to manage without automating triage and issue responses.
If your organization is undergoing digital transformation, modernization, or any new application adoption, service meshes will provide you with critical observability, reliability, and security capabilities while validating and protecting application performance. Use a service mesh, and you’ll get those benefits without writing extra configuration. In today’s hyper-competitive market, engineering teams and business leaders know that businesses must streamline operational processes to focus on what matters most — building great features that your customers love!
Have you considered using service mesh to run modern applications? How did it benefit your business? Share with us on Facebook, Twitter, and LinkedIn. We’d love to hear from you!
MORE ON SERVICE MESH
