Compute & 42 Technical Guides

Comprehensive documentation for deploying, scaling, and optimizing workloads on CloudNexus. Master everything from VM lifecycle management to container orchestration with our structured 42-part guide series.

Compute Updated Dec 2024 v3.4 Series

42 Compute Guides

Each guide is designed to be read sequentially, but can be consumed independently. Covers architecture patterns, CLI commands, Terraform modules, and production best practices.

#01 Beginner

Provisioning Your First NVMe Instance

Step-by-step guide to creating compute instances with high-IOPS storage and optimized kernel parameters.

#02 Beginner

Custom Images & OS Templates

Packaging applications with Cloud-Init, Packer workflows, and snapshot management strategies.

#05 Intermediate

Kubernetes Cluster Bootstrap

Deploying production-ready K8s clusters with CNI plugins, metrics servers, and node auto-scaling policies.

#08 Intermediate

Container Registry & Security Scanning

Private registry setup, vulnerability scanning pipelines, and immutable image tagging conventions.

#12 Intermediate

Global Load Balancer Configuration

L4/L7 load balancing, health checks, connection draining, and traffic splitting strategies.

#15 Advanced

Multi-Region Active-Active Architecture

Cross-region failover, DNS-based routing, and stateless service replication patterns.

#21 Advanced

Compute Optimization & Right-Sizing

Using telemetry data to match workloads with optimal CPU/memory ratios and spot instance strategies.

#28 Advanced

Zero-Trust Compute Environments

Workload identity, mutual TLS, secrets rotation, and hardened baselines for production nodes.

Showing 8 of 42 guides. Use filters or scroll to explore the complete series.

CLI Quick Reference

Manage compute resources programmatically using the cnx compute command group. All commands support JSON output and pagination.

cloudnexus-cli

# Provision a production-optimized instance
cnx compute create --name prod-api-node \
  --type c64.8xlarge \
  --region us-east-1 \
  --image ubuntu-22.04-optimized \
  --storage nvme:200gb \
  --tags env=production,team=platform \
  --auto-scale true

# List running instances with resource metrics
cnx compute list --format table --status running

# Attach GPU workload profile
cnx compute profile set --instance prod-gpu-01 \
  --profile ml-training-a100x4

Flag	Type	Description
`--type`	string	Compute family (c=compute, m=memory, g=GPU, b=baremetal)
`--storage`	string	Format: `type:size` (e.g., nvme:500gb, ssd:1tb)
`--auto-scale`	boolean	Enable intelligent horizontal/vertical scaling policies
`--tags`	string	Key-value pairs for billing, routing, and grouping

Production Best Practices

Compute Sizing Strategy

Start with baseline metrics from staging workloads before provisioning production capacity
Use cnx compute right-size to analyze 30-day utilization patterns
Implement burstable instance families for variable workloads (up to 4x baseline CPU)
Separate stateless compute from stateful storage layers

💡 Pro Tip

Enable Compute Autoscaler v2 in your account settings. It uses ML-based demand forecasting to scale capacity 15 minutes before traffic spikes, reducing cold-start latency by up to 73%.

Networking & Isolation

Always deploy compute nodes within isolated VPC subnets. Use security groups for ingress/egress control and deploy load balancers in public subnets while keeping application tiers private. Enable VPC peering for cross-account resource access without public internet exposure.