Best Practices for Instance Provisioning and Workflow Resilience

Efficient management of cloud infrastructure is critical for maintaining performance, cost control, and operational predictability in live production environments.

Viz Now introduces tools and flexibility to give users greater visibility and control over how compute resources such as EC2 instances are provisioned and consumed.

While Viz Now automates much of the heavy lifting during deployment, cloud infrastructure behaves differently than traditional on-prem setups. Instance availability can fluctuate, quotas may be limited, and GPU resources are often region-specific or subject to supply constraints. Therefore, proactively managing your capacity strategy is essential.

In this section, we’ll walk through key strategies to:

  • Plan and reserve compute capacity for mission-critical workflows using default Viz Now Targeted Capacity Reservations

  • Understand Guided Retry and Fail-Fast Deployment to avoid mid-process launch failures

  • Interpret Visual Indicators to assess deployment capacity status in real time

  • Leverage BYOC (Bring Your Own Capacity) to manually assign long-term reserved capacity

  • Change EC2 Instance Types dynamically to adapt to availability and cost constraints

  • Switch Regions or Availability Zones to bypass regional shortages

  • Collaborate with AWS TAMs to understand capacity trends and proactively plan deployments.

Capacity Management Strategies for Resilient Cloud Deployments:

These best practices are designed to help you make informed decisions when deploying Viz Now spaces and ensure your production is not impacted by unexpected resource shortages.