AWS Marketplace Deployment

Deploy AxonFlow in your AWS account using the AWS Marketplace Container product. This guide walks you through the end-to-end deployment process using AWS CloudFormation.

Overview

AxonFlow In-VPC Deployment runs entirely within your AWS account, providing:

• Sub-10ms P95 latency (vs ~100ms for public SaaS)
• Complete data privacy (no data leaves your VPC)
• Node-based licensing (Professional: 10 nodes, Enterprise: 50 nodes, Enterprise Plus: unlimited)
• Production-ready infrastructure (Multi-AZ RDS, ECS Fargate, ALB)

Deployment Time: 15-25 minutes

Prerequisites

1. AWS Account Requirements

• AWS Account with appropriate permissions
• VPC with at least 2 private subnets across different AZs
• IAM Permissions:
  • CloudFormation: Create/Update/Delete stacks
  • ECS: Create services, task definitions
  • RDS: Create DB instances
  • EC2: Create security groups, load balancers
  • IAM: Create roles for ECS tasks

2. Required Information

Before deployment, gather:

Item	Example	Where to Find
VPC ID	vpc-0123abcd	VPC Console
Subnet IDs	subnet-abc123, subnet-def456	VPC Console → Subnets (2+ private subnets)
License Key	AXON-PRO-12345-ABCDE	Provided by AxonFlow sales after purchase
ACM Certificate ARN (optional)	arn:aws:acm:...	ACM Console (for HTTPS)

3. License Tiers

Tier	Max Nodes	Annual Price	Best For
Professional	10	$240K	Small teams, 100K requests/day
Enterprise	50	$720K	Growing companies, 500K requests/day
Enterprise Plus	Unlimited	Custom	Large enterprises, millions of requests/day

Note: Nodes = total Agent + Orchestrator containers

Deployment Steps

Option 1: AWS Console (Recommended for First Deployment)

Step 1: Subscribe to AxonFlow on AWS Marketplace

1. Go to AWS Marketplace
2. Search for "AxonFlow"
3. Click Continue to Subscribe
4. Review pricing and terms
5. Click Subscribe
6. Wait for subscription confirmation (~2 minutes)
7. Click Continue to Configuration

Step 2: Configure Deployment

1. Fulfillment Option: Container
2. Software Version: Latest (1.0.0)
3. Region: Select your region (e.g., eu-central-1)
4. Click Continue to Launch

Step 3: Launch CloudFormation Stack

1. Choose Action: Launch CloudFormation
2. Click Launch
3. You'll be redirected to CloudFormation console

Step 4: Configure Stack Parameters

Stack Details:

• Stack Name: axonflow-production

Network Configuration:

• VPC ID: Select your VPC
• Subnet IDs: Select 2+ private subnets (different AZs)

Capacity Configuration:

• Agent Desired Count: 2 (start with 2, scale later)
• Orchestrator Desired Count: 2 (start with 2, scale later)
• Agent Task CPU: 1024 (1 vCPU)
• Agent Task Memory: 2048 (2 GB)
• Orchestrator Task CPU: 1024 (1 vCPU)
• Orchestrator Task Memory: 2048 (2 GB)

Database Configuration:

• DB Instance Class: db.t3.medium (2 vCPU, 4 GB) - start small
• DB Username: axonflow_admin (recommended)
• DB Password: Generate secure password (20+ characters)
• DB Allocated Storage: 100 GB (gp3 SSD)

HTTPS Configuration (Optional):

• Certificate ARN: Paste ACM certificate ARN or leave blank for HTTP

License Configuration:

• License Tier: Select Professional, Enterprise, or EnterprisePlus

Step 5: Deploy

1. Acknowledge IAM resource creation
2. Click Create Stack
3. Wait 15-25 minutes for deployment

Stack Events to Watch:

• CREATE_IN_PROGRESS: Stack creation started
• AgentService CREATE_IN_PROGRESS: Agent tasks starting (~10 min)
• OrchestratorService CREATE_IN_PROGRESS: Orchestrator tasks starting (~12 min)
• CREATE_COMPLETE: Deployment successful

Option 2: AWS CLI (For Automation)

Step 1: Download CloudFormation Template

Download from AWS Marketplace or AxonFlow:

curl -O https://axonflow-public-assets.s3.amazonaws.com/cloudformation/ecs-fargate.yaml

Step 2: Create Parameters File

Create parameters.json:

[
  {
    "ParameterKey": "VPC",
    "ParameterValue": "vpc-0123abcd"
  },
  {
    "ParameterKey": "Subnets",
    "ParameterValue": "subnet-abc123,subnet-def456"
  },
  {
    "ParameterKey": "AgentDesiredCount",
    "ParameterValue": "2"
  },
  {
    "ParameterKey": "OrchestratorDesiredCount",
    "ParameterValue": "2"
  },
  {
    "ParameterKey": "DBInstanceClass",
    "ParameterValue": "db.t3.medium"
  },
  {
    "ParameterKey": "DBUsername",
    "ParameterValue": "axonflow_admin"
  },
  {
    "ParameterKey": "DBPassword",
    "ParameterValue": "YOUR_SECURE_PASSWORD_HERE"
  },
  {
    "ParameterKey": "LicenseTier",
    "ParameterValue": "Professional"
  }
]

Step 3: Deploy Stack

aws cloudformation create-stack \
  --stack-name axonflow-production \
  --template-body file://ecs-fargate.yaml \
  --parameters file://parameters.json \
  --capabilities CAPABILITY_IAM \
  --region eu-central-1

Step 4: Monitor Deployment

aws cloudformation describe-stacks \
  --stack-name axonflow-production \
  --region eu-central-1 \
  --query "Stacks[0].StackStatus"

# Wait for CREATE_COMPLETE
aws cloudformation wait stack-create-complete \
  --stack-name axonflow-production \
  --region eu-central-1

Post-Deployment Configuration

1. Retrieve Stack Outputs

Get the Agent endpoint URL:

aws cloudformation describe-stacks \
  --stack-name axonflow-production \
  --region eu-central-1 \
  --query "Stacks[0].Outputs"

Key Outputs:

• AgentEndpoint: Internal load balancer URL (e.g., http://internal-axonflow-alb-123.eu-central-1.elb.amazonaws.com)
• DatabaseEndpoint: RDS endpoint (internal use)
• TierMaxNodes: Node limit for your tier (10, 50, or -1)
• TierAnnualPrice: Annual license price

2. Health Check

Verify Agent is healthy:

curl -s http://internal-axonflow-alb-123.eu-central-1.elb.amazonaws.com/health | jq

Expected response:

{
  "status": "healthy",
  "agent_version": "1.0.0",
  "license": {
    "tier": "Professional",
    "maxNodes": 10,
    "valid": true
  }
}

3. Configure Your Application

TypeScript/Node.js:

import { AxonFlow } from '@axonflow/sdk';

const axonflow = new AxonFlow({
  apiKey: 'your-client-id',  // Provided by AxonFlow
  endpoint: 'http://internal-axonflow-alb-123.eu-central-1.elb.amazonaws.com'
});

Go:

import "github.com/getaxonflow/axonflow-sdk-go"

client := axonflow.NewClientSimple(
    "http://internal-axonflow-alb-123.eu-central-1.elb.amazonaws.com",
    "your-client-id",
    "your-client-secret",
)

4. Test End-to-End

Test a simple query:

curl -X POST http://internal-axonflow-alb-123.eu-central-1.elb.amazonaws.com/v1/execute \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer your-client-id" \
  -d '{
    "prompt": "What is 2+2?",
    "mode": "chat"
  }'

Expected: {"success": true, "data": "4", ...}

Scaling

Auto-Scaling (Automatic)

AxonFlow automatically scales based on CPU utilization:

• Target: 70% CPU
• Scale Out: Add tasks when CPU > 70%
• Scale In: Remove tasks when CPU < 50%
• Limits: Min 1, Max 50 tasks per component

Manual Scaling

Update task counts via CloudFormation:

aws cloudformation update-stack \
  --stack-name axonflow-production \
  --use-previous-template \
  --parameters \
    ParameterKey=AgentDesiredCount,ParameterValue=5 \
    ParameterKey=OrchestratorDesiredCount,ParameterValue=5 \
  --region eu-central-1

Scaling Guidelines:

Request Volume	Recommended Agent	Recommended Orchestrator
<10K req/day	2	2
10K-100K req/day	3-5	3-5
100K-500K req/day	10-20	10-15
500K+ req/day	20-50	15-30

Monitoring

CloudWatch Metrics

AxonFlow publishes metrics to CloudWatch:

Agent Metrics:

• CPUUtilization: Agent CPU usage
• MemoryUtilization: Agent memory usage
• TargetResponseTime: Agent latency (P50, P95, P99)
• RequestCount: Total requests

Orchestrator Metrics:

• CPUUtilization: Orchestrator CPU usage
• MemoryUtilization: Orchestrator memory usage
• TaskCount: Active tasks

RDS Metrics:

• DatabaseConnections: Active DB connections
• FreeStorageSpace: Remaining storage

CloudWatch Logs

View logs:

# Agent logs
aws logs tail /ecs/axonflow-agent --follow --region eu-central-1

# Orchestrator logs
aws logs tail /ecs/axonflow-orchestrator --follow --region eu-central-1

Create Alarms

High CPU Alarm:

aws cloudwatch put-metric-alarm \
  --alarm-name axonflow-agent-high-cpu \
  --alarm-description "Agent CPU > 80%" \
  --metric-name CPUUtilization \
  --namespace AWS/ECS \
  --statistic Average \
  --period 300 \
  --threshold 80 \
  --comparison-operator GreaterThanThreshold \
  --evaluation-periods 2

Security Best Practices

1. Network Isolation

• Deploy in private subnets (no internet gateway)
• Use internal ALB (not internet-facing)
• Control access with security groups

2. Database Security

• Encryption at rest: Enabled by default
• Encryption in transit: SSL/TLS required
• Automated backups: 7-day retention
• Rotate password: Every 90 days via Secrets Manager

3. IAM Roles

• Least privilege: ECS task roles have minimal permissions
• No API keys in containers: Use AWS Secrets Manager

4. License Key Protection

Store license key in AWS Secrets Manager:

aws secretsmanager create-secret \
  --name axonflow/license-key \
  --secret-string "AXON-PRO-12345-ABCDE-67890" \
  --region eu-central-1

Troubleshooting

Issue 1: Stack Creation Failed

Symptom: CloudFormation rollback

Common Causes:

• Insufficient IAM permissions
• VPC subnets in same AZ
• Invalid license key

Solution:

1. Check CloudFormation Events tab for error
2. Verify VPC has 2+ subnets in different AZs
3. Validate IAM permissions
4. Delete failed stack and retry

Issue 2: Health Check Failing

Symptom: curl /health returns 503 or times out

Solution:

# Check ECS service status
aws ecs describe-services \
  --cluster axonflow-cluster \
  --services axonflow-agent-service \
  --region eu-central-1

# Check task health
aws ecs list-tasks \
  --cluster axonflow-cluster \
  --service-name axonflow-agent-service \
  --region eu-central-1

# View task logs
aws logs tail /ecs/axonflow-agent --since 10m --region eu-central-1

Issue 3: License Invalid

Symptom: Health check shows "license": { "valid": false }

Solution:

1. Verify license key format: AXON-{TIER}-{5 digits}-{5 chars}-{5 chars}
2. Check expiration date
3. Contact AxonFlow support for license renewal
4. Update license via Secrets Manager or environment variable

Issue 4: High Latency

Symptom: Latency > 10ms P95

Causes:

• Insufficient capacity (CPU/memory)
• Database performance issues
• Cross-AZ traffic

Solution:

1. Scale up task count
2. Increase task CPU/memory
3. Check RDS performance metrics
4. Use VPC endpoint for SDK connections

Cost Estimation

Professional Tier (10 nodes, ~50K requests/day)

Monthly AWS Costs:

Service	Configuration	Monthly Cost
ECS Fargate	4 tasks × 1 vCPU × 2GB × 730h	$213
RDS	db.t3.medium, 100GB, Multi-AZ	$182
ALB	Internal, ~5GB processed	$25
Data Transfer	Intra-VPC (free)	$0
CloudWatch	Logs 10GB, metrics	$10
Backup Storage	7 days × 100GB	$19
Total AWS		~$449/month
License Fee	Professional tier annual	$20K/month
Grand Total		~$20,449/month

Enterprise Tier (50 nodes, ~500K requests/day)

Monthly AWS Costs:

Service	Configuration	Monthly Cost
ECS Fargate	20 tasks × 1 vCPU × 2GB × 730h	$1,065
RDS	db.m5.xlarge, 500GB, Multi-AZ	$320
ALB	Internal, ~50GB processed	$28
CloudWatch	Logs 100GB, metrics	$30
Backup Storage	7 days × 500GB	$95
Total AWS		~$1,538/month
License Fee	Enterprise tier annual	$60K/month
Grand Total		~$61,538/month

Cost Optimization Tips:

• Use Reserved Instances for Fargate (save 20-40%)
• Optimize RDS instance size based on actual load
• Enable RDS storage autoscaling
• Use CloudWatch Logs retention policies (7-30 days)

Upgrading

Upgrade AxonFlow Version

AWS Marketplace will notify you of new versions. To upgrade:

1. Subscribe to new version in AWS Marketplace
2. Update CloudFormation stack with new container images
3. ECS will perform rolling update (zero downtime)

aws cloudformation update-stack \
  --stack-name axonflow-production \
  --use-previous-template \
  --parameters \
    ParameterKey=AgentImageTag,ParameterValue=1.1.0 \
    ParameterKey=OrchestratorImageTag,ParameterValue=1.1.0 \
  --region eu-central-1

Upgrade License Tier

Contact AxonFlow sales to upgrade from Professional → Enterprise → Enterprise Plus.

After receiving new license key:

# Update license in Secrets Manager
aws secretsmanager update-secret \
  --secret-id axonflow/license-key \
  --secret-string "AXON-ENT-67890-FGHIJ-KLMNO" \
  --region eu-central-1

# Restart ECS tasks to apply new license
aws ecs update-service \
  --cluster axonflow-cluster \
  --service axonflow-agent-service \
  --force-new-deployment \
  --region eu-central-1

Support

Having deployment issues?

• Email: [email protected]
• Documentation: https://docs.getaxonflow.com
• Slack: Join our community

Include in your support request:

• CloudFormation stack name
• AWS region
• Stack events (from CloudFormation console)
• ECS task logs (if applicable)
• License tier

Next Steps

• Read the TypeScript SDK Guide to integrate AxonFlow
• Read the Go SDK Guide for Go applications
• Learn about Authentication and security best practices
• Explore MCP Connectors for external data integration

Additional Resources

• AWS Marketplace Listing: AxonFlow on AWS Marketplace
• CloudFormation Template: Available in AWS Marketplace
• Architecture Diagram: See Architecture Guide
• Pricing Details: https://getaxonflow.com/enterprise-pricing

---

Last Updated: October 28, 2025