How Intuit improves security, latency, and development velocity with a service mesh
In this sponsored episode of the podcast, we talk with Anil Attuluri, principal software engineer, and Yasen Simeonov, senior product manager, both of Intuit, about how their engineering organization uses a service mesh to solve problems, letting their engineers stay focused on writing business logic. Along the way, we discuss how the service mesh keeps all the financial data secure, how it moves network traffic to where it needs to go, and the open source software they’ve written on top of the mesh.
At an SaaS company like Intuit that has hundreds of services spread out across multiple products, maintaining development velocity at scale means baking some of the features that every service needs into the architecture of their systems. That’s where a service mesh comes in. It automatically adds features like observability, traffic management, and security to every service in the network without adding any code.
In this sponsored episode of the podcast, we talk with Anil Attuluri, principal software engineer, and Yasen Simeonov, senior product manager, both of Intuit, about how their engineering organization uses a service mesh to solve problems, letting their engineers stay focused on writing business logic. Along the way, we discuss how the service mesh keeps all the financial data secure, how it moves network traffic to where it needs to go, and the open source software they’ve written on top of the mesh.
Episode notes:
For those looking to get the same service mesh capabilities as Intuit, check out Istio, a Cloud Native Computing Foundation project.
In order to provide a better security posture for their products, each business case operates on a discrete network. But much of the Istio service mesh needs to discover services across all products. Enter Admiral, their open-sourced solution.
When Intuit deploys a new service version, they can progressively scale the amount of traffic that hits it instead of the old version using Argo Rollouts. It’s better to find a bug in production on 1% of requests than 100%.
If you want to learn more about what Intuit engineering is doing, check out their blog.
Congrats to Great Question badge winner,
, for asking