Skip to content

MODULAR-APPS.md

Latest commit

 

History

History
305 lines (222 loc) · 7.34 KB

MODULAR-APPS.md

File metadata and controls

305 lines (222 loc) · 7.34 KB

Extending MicroTraderX with Additional Applications (Conceptual)

ConfigHub Documentation: For Sets, Filters, WHERE queries, and cross-space discovery patterns, see docs.confighub.com. This document shows conceptual extension patterns for MicroTraderX.

Experimental approach to adding auxiliary applications using separate ConfigHub spaces.

Note: This document presents a conceptual architecture pattern. These extensions (stages 8-9) are not implemented in MicroTraderX. For production-ready implementations of these concepts, see:

MicroTraderX (stages 1-7) remains a focused tutorial. This document shows how you could extend it using ConfigHub's modular architecture.

Problem Statement

Traditional approaches require modifying deployment configurations, adding dependencies, coordinating releases, managing version compatibility, and creating custom integrations.

ConfigHub approach: Applications in separate spaces discover resources via WHERE queries.

Base Configuration Structure

Stages 1-7

traderx-base/
├── reference-data
└── trade-service

traderx-prod-us/
traderx-prod-eu/
traderx-prod-asia/

Extension Pattern

Additional applications in separate spaces:

cost-optimizer/
├── analyzer
├── recommendations
└── dashboard

drift-detector/
├── detector
└── corrector

Stage 8: Cost Optimization Application

Label Base Resources

# Conceptual example - shows the pattern
# For correct --patch syntax, see TESTING.md

for region in us eu asia; do
  # In practice, use: echo '{...}' | cub unit update ... --patch --from-stdin
  cub unit update trade-service --space traderx-prod-$region \
    --patch '{"Labels":{"app":"microtraderx","cost-optimizable":"true"}}'
  cub unit update reference-data --space traderx-prod-$region \
    --patch '{"Labels":{"app":"microtraderx","cost-optimizable":"true"}}'
done

Deploy Analyzer

cub space create cost-optimizer

cat > cost-analyzer.yaml << 'EOF'
apiVersion: batch/v1
kind: CronJob
metadata:
  name: cost-analyzer
spec:
  schedule: "0 * * * *"
  jobTemplate:
    spec:
      template:
        spec:
          containers:
          - name: analyzer
            image: cost-optimizer:latest
            command: ["./discover-and-analyze"]
EOF

cub unit create analyzer --space cost-optimizer --data cost-analyzer.yaml
cub unit apply analyzer --space cost-optimizer

Discovery Query

# Conceptual query (Labels.field notation may not be supported)
# In practice, use simpler queries or label-based discovery

# Example: Find all units in production spaces
cub unit list --space "*" \
  --where "Space.Slug LIKE '%prod%' AND Slug LIKE '%service'"

# Or query by specific unit names across all spaces
cub unit list --space "*" \
  --where "Slug = 'trade-service'"

Conceptually returns units like:

  • traderx-prod-us/trade-service
  • traderx-prod-eu/trade-service
  • traderx-prod-asia/trade-service

For production implementation, see devops-examples/cost-optimizer which uses tested WHERE patterns.

Apply Optimization

cub run set-replicas --replicas 1 \
  --space traderx-prod-eu \
  --unit trade-service

cub unit create optimization-2024-01-15 --space cost-optimizer \
  --data '{"saved":"$500/month","region":"eu","action":"scale-down"}'

Result: Base MicroTraderX configuration unchanged.

Stage 9: Drift Detection Application

cub space create drift-detector

cub unit list --space "traderx-prod-*" \
  --columns Slug,Data,LiveState

cub changeset create fix-drift
cub unit apply trade-service --space traderx-prod-us
cub changeset apply fix-drift

Integration Patterns

Discovery via WHERE Queries

# Conceptual example - complex NOT LIKE patterns may not be supported
# In practice, use simpler queries

# Find all units with "service" in the name
cub unit list --space "*" \
  --where "Slug LIKE '%service'"

# Find units in specific spaces
cub unit list --space "traderx-prod-*" \
  --where "Slug != ''"

# For replica filtering, query in application code after fetching
# See devops-examples/cost-optimizer for production patterns

Links for Relationships

cub link create cost-monitoring \
  traderx-prod-eu/trade-service \
  cost-optimizer/analyzer

Sets for Grouping

cub set create business-critical
cub set add-unit business-critical \
  traderx-prod-us/trade-service \
  traderx-prod-eu/trade-service \
  cost-optimizer/analyzer

Filters as Saved Queries

cub filter create trading-services Unit \
  --where-field "Labels.app = 'microtraderx' AND
                 Slug LIKE '%trade-service%'" \
  --space traderx-base

cub unit list --filter traderx-base/trading-services --space "*"

Complete Structure

ConfigHub spaces after extension:

├── traderx-base/
├── traderx-prod-us/
├── traderx-prod-eu/
├── traderx-prod-asia/
├── cost-optimizer/
└── drift-detector/

Integration method:

  • cost-optimizer discovers traderx via WHERE queries
  • drift-detector monitors traderx LiveState
  • No hardcoded dependencies between spaces

Comparison with Traditional Approaches

Traditional tools:

  • Helm: Requires chart modifications, dependency declarations
  • Kustomize: Requires patching overlays
  • GitOps: Requires repository modifications, coordination
  • Terraform: Cannot query across workspaces

ConfigHub approach:

  • Spaces provide module boundaries
  • WHERE queries enable discovery without hardcoding
  • Base application configuration unchanged
  • Applications discover each other at runtime

Application Template

#!/bin/bash

APP_NAME="my-devops-app"
cub space create $APP_NAME

cat > app.yaml << EOF
metadata:
  labels:
    app: $APP_NAME
    discoverable: true
    capabilities: "monitoring,analysis"
EOF

cub unit create app --space $APP_NAME --data app.yaml

TARGETS=$(cub unit list --space "*" \
  --where "Labels.app = 'microtraderx'")

for target in $TARGETS; do
  echo "Found: $target"
  # Application logic
done

cub unit create results-$(date +%Y%m%d) \
  --space $APP_NAME \
  --data results.json

Example Operations

cub unit list --space drift-detector --where "status = 'drifted'"
cub unit get latest-analysis --space cost-optimizer

for region in eu asia; do
  cub run set-replicas --replicas 1 \
    --space traderx-prod-$region \
    --unit trade-service
done

Summary

Pattern demonstrated:

  • Base application (2 services, 3 regions)
  • Extension applications (separate spaces)
  • Discovery via WHERE queries
  • Loose coupling (no hardcoding)

Implementation steps:

  1. Create separate space
  2. Discover target resources
  3. Implement application logic
  4. Store results as units

Documentation References