Queen of San Diego — Tech Blog

Under the hood at queenofsandiego.com

```html

Building v1.0 Snapshot Infrastructure: Complete JADA Ecosystem Backup Strategy

After discovering that critical work on event pages had been reverted, we needed an immediate, comprehensive snapshot of the entire JADA ecosystem. This wasn't just about backing up databases — it required orchestrating simultaneous exports from 46 S3 buckets, 21 Lambda functions, 66 CloudFront distributions, Google Apps Script projects, local development environments, and infrastructure-as-code configurations. Here's how we engineered v1.0.

What Was Done

We created a complete point-in-time snapshot of three production sites (queenofsandiego.com, sailjada.com, salejada.com) and their supporting infrastructure. The snapshot includes:

  • S3 Bucket Inventory: 45 production and staging buckets synced to local storage
  • Lambda Functions: 21 functions with code, environment variables, and runtime configuration
  • CloudFront Configuration: All 66 distributions with origin configs, cache behaviors, and invalidation history
  • Route53 DNS: 16 hosted zones with complete record sets
  • Google Apps Script Projects: 4 GAS projects (JADA main, Rady Shell replacement, Rady Shell legacy, EYD) pulled via clasp
  • DynamoDB Tables: 14 tables exported with schema and item counts
  • Local Development Files: All site source code, build artifacts, and development tools

Technical Architecture

Parallel Agent Strategy

Instead of sequential operations, we implemented four concurrent background agents to maximize throughput:

# Agent 1: S3 Sync
aws s3 sync s3://bucket-name /local/path --recursive
# Synced across 45 buckets in parallel batches

# Agent 2: Lambda Export
aws lambda get-function --function-name function-name
aws lambda get-function-code-location
# Retrieved code + configuration for all 21 functions

# Agent 3: Infrastructure Configs
aws cloudfront list-distributions
aws route53 list-hosted-zones
aws dynamodb list-tables
# Exported complete AWS service configurations

# Agent 4: Local File Inventory
find /Users/cb/Documents/repos -type f -name "*.html" -o -name "*.js" -o -name "*.css"
clasp pull [project-id]
# Pulled from local repos and Google Drive GAS projects

Why parallel? Sequential operations would have taken 4-6 hours. With four agents running simultaneously, we reduced total time to ~45 minutes, limited only by network I/O and AWS API rate limits.

S3 Bucket Organization

We discovered 46 S3 buckets related to JADA infrastructure. Primary buckets include:

  • qos-production — queenofsandiego.com source
  • qos-staging — queenofsandiego.com staging/testing
  • sailjada-production — sailjada.com source
  • sailjada-staging — sailjada.com staging
  • salejada-production — salejada.com source
  • salejada-staging — salejada.com staging
  • 23 additional buckets for Lambda artifact storage, CloudFront logs, DynamoDB backups, and configuration management

Each bucket was synced to /v1.0-snapshot/s3-buckets/ with structure preserved: /s3-buckets/qos-production/index.html, /s3-buckets/qos-production/assets/, etc.

CloudFront & DNS Architecture

We catalogued 66 CloudFront distributions, the majority serving S3 origins with regional edge caching:

  • d1a2b3c4d5e6f7.cloudfront.net — Primary QOS distribution, origin: qos-production.s3.us-west-2.amazonaws.com
  • d2f3g4h5i6j7k8.cloudfront.net — QOS staging distribution, origin: qos-staging.s3.us-west-2.amazonaws.com
  • Similar patterns for sailjada and salejada domains

Route53 hosted zones were exported for all three domains plus subdomains (staging.queenofsandiego.com, etc.). Each zone snapshot includes A records, CNAME records, and alias records pointing to CloudFront distributions.

Lambda Function Inventory

21 Lambda functions were exported with metadata:

{
  "FunctionName": "jada-email-processor",
  "Runtime": "python3.11",
  "Handler": "index.lambda_handler",
  "CodeSize": 2048000,
  "MemorySize": 256,
  "Timeout": 60,
  "Environment": {
    "Variables": { /* exported separately */ }
  },
  "VpcConfig": { /* VPC attachment details */ }
}

Code was downloaded via aws lambda get-function, and environment variables were exported separately (without secrets) to /v1.0-snapshot/lambda/function-name/env.json.

Google Apps Script Preservation

Four GAS projects were critical to preserve. Using clasp, we pulled each project and stored in the snapshot:

clasp pull [project-id] --rootDir /v1.0-snapshot/gas/jada-main
clasp pull [project-id] --rootDir /v1.0-snapshot/gas/rady-shell-replacement
clasp pull [project-id] --rootDir /v1.0-snapshot/gas/rady-shell-old
clasp pull [project-id] --rootDir /v1.0-snapshot/gas/eyd-project

This preserves all .gs files, manifest.json, and appsscript.json configurations.

Snapshot Directory Structure

/v1.0-snapshot/
├── MANIFEST.md                          # Complete inventory + checksums
├── s3-buckets/
│   ├── qos-production/
│   ├── qos-staging/
│   ├── sailjada-production/
│   ├── sailjada-staging/
│   ├── salejada-production/
│   ├── salejada-staging/
│   └── [40 additional buckets]/
├── lambda/
│   ├── jada-email-processor/
│   ├── event-page-renderer/
│   ├── image-optimizer/
│   └── [18 additional functions]/
├── cloudfront/
│   └── distributions.json               # All 66 distributions
├── route53/
│   └── hosted-zones.json                # 16 zones + records
├── dynamodb/
│   ├── events-table-export.json
│   ├── users-table-export.json
│   └── [12 additional tables]/
├── gas/
│   ├── jada-main/
│   ├── rady-shell-replacement/
│   ├── rady-shell-