Multiple Environment Management at Vymo with Facets

How Vymo Leveraged Facets to Accelerate Kubernetes Migration, Enable Multi-Cloud Scalability, and Achieve 200+ Monthly Releases.

Strategic Value Delivered by Facets

Before

1. Complexity of managing multiple production environments for customers from different geographies

2. A lean Ops team was tasked with transformations like Kubernetes migration and optimizing Azure spend

3. Assumptions in application code that hindered the adoption of standard scaling strategies, leading to over-provisioning and cloud wastage

After

1. Managing more than 15 environments across multiple clouds pushing over 200 releases a month

2. Accelerated Kubernetes migration and well-automated infrastructure within weeks of implementation

3. Co-developed custom scaling strategies on top of the Platform saving 12% of cloud cost

Challenges Faced by Vymo

1. Increased complexity due to multiplicity of environments: Supporting BFSI clients meant managing custom workflows and maintaining multiple production environments across geographies. Each client often demanded unique deployment timelines and configurations, adding to operational complexity.

• Drifts: Maintaining drifts across multiple regions without a centralized management solution created more work for the Ops teams.

• Multi-cloud setups: While primarily on Azure, Vymo wanted cloud flexibility to honor region or customer-specific preferences. Hence, they maintained a cloud-agnostic codebase, however, taking the same approach in automation wasn’t trivial.

• Ops heavy Releases: As Vymo provides tailored customization for customers, release frequency and artifacts vary across environments; leading to operational overheads in releases.

2. Limited Ops bandwidth and long roadmap: The Ops team was lean making it challenging to undertake large-scale projects like Kubernetes migration and 100% automation to manage multiple environments. The manual nature of releases and maintenance required constant intervention from the operations team, leaving little bandwidth for strategic improvements.

3. Poor Cloud Utilization: It was critical to manage cloud costs while scaling environments for performance and reliability, especially with the introduction of Kubernetes and new deployment strategies.

• Lack of Elasticity: Vymo's infrastructure relied heavily on virtual machines (VMs) with Ansible-based configuration management. While functional, this setup lacked elasticity, scalability, and developer-centric workflows, which limited operational efficiency. Also, many services were not built keeping usual autoscaling in mind. Thus, they were over-provisioned leading to wastage. This also prevented the use of a standard spot instance management strategy without sacrificing reliability.

“When I joined Vymo, the systems were functional but rigid. Moving to Kubernetes offered the promise of elasticity and efficiency, but we needed a partner who could help us bridge the gap without disrupting our workflows”

• Cost of observability: The Ops team lacked the bandwidth to fine-tune open-source solutions which increased the cost of observability infrastructure.

How Facets Transformed Vymo's DevOps

Drift-free multiple environments: Facets made environment management easy, self-service and enabled consistent deployment practices across multiple geographies and data centers.

Accelerating DevOps Roadmap: Facets accelerated the DevOps transformation projects

• Kubernetes Migration: Vymo adopted a parallel setup, running both VMs and Kubernetes during migration.
• Well-architected Observability: Developers were transitioned to Kubernetes-native live logs enriched by Facets' platform.

Optimized Cloud posture: Facets and Vymo's team collaborated to add custom scaling strategies on top of the platform to meet the application elasticity assumptions.

Key Results

1. Increased Release Frequency: Vymo performs over 200 releases per month with central tracking and management.

2. Cost Savings: Vymo achieved 12% cloud cost savings with migration to Kubernetes and custom scaling strategies developed over the platform

3. Faster Onboarding: Vymo is now able to onboard environments and get them up and running faster.

4. Improved Uptime: A combination of Kubernetes migration, custom scaling strategies, and proactive alerting ensured that Vymo maintained a very high availability for customers. ‍ “No amount of testing and validation can guarantee zero downtime, but we’ve avoided many potential issues due to Facets,” says Pruthvi.

5. Performance Gains: Kubernetes clusters were optimized for nuanced geographies like Japan, where uptime and precision are critical. Fine-tuned configurations ensured consistent performance across multiple environments.

What's next for Vymo

Vymo is looking forward to expanding support for more clouds and on-premise solutions through Facets to expand to more regions and customer segments. They aim to do this cost-optimally while minimizing distractions to the broader organization. They are also in the process of adopting the next level of process standards.

"It was the ease of adoption and the founders knew what they were doing when they started Facets," noted Pruthvi Narapareddy, Director of Engineering at Vymo.

"