feat(providers/azure/services): add Cache for Redis support

[cristim]

Summary

AWS MemoryDB is a multi-AZ in-memory database service; Azure lacks a direct equivalent in the current CUDly provider implementation. Azure offers managed in-memory caching via Azure Cache for Redis and in-memory databases via other services, but no MemoryDB-equivalent is integrated.

What This Issue Is

Adding Azure in-memory database support (Azure Cache for Redis or similar managed service):

• Discovery of managed Redis instances and equivalents
• Performance tier analysis and rightsizing recommendations
• Commitment-based discount options where available

What This Issue Isn't

• Standalone Redis deployments on VMs (covered under compute)
• Replication architecture decisions
• Multi-region failover strategies

Suggested Approach

Reference AWS MemoryDB implementation in providers/aws/services/ for the service structure. Map to Azure Cache for Redis using Azure Compute Management API, including tier enumeration, utilization metrics, and commitment pricing from the Azure Pricing API.

Acceptance Criteria

• Azure Cache for Redis instances are discovered
• Tier utilization and performance metrics collected
• Rightsizing recommendations generated (scale up/down, tier changes)
• Commitment-based pricing integrated
• Integration tests verify resource discovery and cost analysis

Refs #473

[cristim]

Verification: shipped in PR #620 (SHA 78d3273, merged 2026-05-22). providers/azure/services/managedredis/client.go added with full Azure Cache for Redis reserved-capacity client, 592 lines including GetRecommendations. Wired into azure provider.go and services.go. Confirmed on origin/feat/multicloud-web-frontend.