Improve Deckard integration with Redis Clusters

[Copilot]

This PR implements Redis Cluster support for Deckard while maintaining full backward compatibility with single-node Redis deployments.

Problem

Deckard's Lua scripts access multiple keys that may hash to different slots in a Redis Cluster, causing script execution to fail. Redis Cluster requires all keys accessed by a Lua script to be on the same hash slot for atomic operations.

Solution

Implemented a hash tag strategy that ensures all queue-related keys land on the same Redis Cluster slot:

Before (Single-node only):

• deckard:queue:myqueue
• deckard:queue:myqueue:tmp
• deckard:queue:myqueue:lock_ack

After (Cluster-compatible):

• deckard:queue:{myqueue}
• deckard:queue:{myqueue}:tmp
• deckard:queue:{myqueue}:lock_ack

The {myqueue} hash tag ensures atomic operations across all related keys.

Changes

Configuration Support

• Added redis.cluster.mode and redis.cluster.addresses configuration options
• Support for both redis.Client and redis.ClusterClient via unified interface
• Environment variable support: DECKARD_REDIS_CLUSTER_MODE, DECKARD_REDIS_CLUSTER_ADDRESSES

Key Naming Strategy

• Modified key generation methods to use hash tags when cluster mode is enabled
• Maintains backward compatibility - single-node deployments unchanged
• Applied to all Redis operations: queue pools, locks, and general cache keys

Testing Infrastructure

• Redis Cluster Docker Compose setup with 6 nodes (3 masters, 3 replicas)
• Comprehensive integration tests validating Lua script compatibility
• Unit tests for both single-node and cluster key generation
• Automated cluster setup script for local development

Documentation

• Complete Redis Cluster configuration guide
• Migration instructions from single-node to cluster
• Docker Compose examples and troubleshooting guide
• Performance considerations and best practices

Configuration Examples

Single-node (default, unchanged):

cache:
  type: REDIS
redis:
  address: localhost
  port: 6379

Redis Cluster (new):

cache:
  type: REDIS
redis:
  cluster:
    mode: true
    addresses:
      - redis-node-1:6379
      - redis-node-2:6379
      - redis-node-3:6379

Backward Compatibility

• ✅ Zero breaking changes to existing API
• ✅ Single-node Redis deployments work unchanged
• ✅ All existing tests pass
• ✅ Default configuration remains single-node mode
• ✅ Existing key naming preserved for single-node mode

Testing

Run cluster integration tests:

./scripts/setup-redis-cluster.sh
export REDIS_CLUSTER_TEST=1
go test -v ./internal/queue/cache/ -run Cluster

Fixes #13.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[lucasoares]

@copilot re-run checks so I can check if tests pass

[Copilot]

@copilot re-run checks so I can check if tests pass

Checks re-triggered in commit 6245e08. The CI/CD pipeline should now run the full test suite for the Redis Cluster implementation.

[cezar-tech]

can you make it easier to assimilate by starting the node count from 0 so that it conforms to the ports final digits ?

[cezar-tech]

since you are using {myqueue} to refer to a hash tag, this templated name should be improved to be more mnemonic dont you think so ?
call it {clusterhash} for example.

and change this terminology in all the documentations you have made.

[cezar-tech]

this is repeated twice, maybe transform it into a method ?