Roadmap

Completed

• Core time-series storage engine
• WAL with batch writer and partitioned storage
• V6 columnar format (single-file parts with footer index)
• Adaptive compression (Gorilla, Delta, Dictionary, Bool + Snappy)
• 3-tier group-aware TieredStorage (Group → Device → Partition)
• 64-shard MemoryStore for hot data
• Event-driven FlushWorkerPool
• Background CompactionWorker
• Multi-level aggregation pipeline (1h/1d/1M/1y cascading)
• Device-based sharding with GroupAutoAssigner
• Adaptive hashing (Rendezvous + Consistent)
• NATS/Redis/Kafka/Memory message queue support
• gRPC inter-service communication with connection pooling
• Scatter-gather QueryCoordinator
• API key authentication (X-API-Key, Bearer, ApiKey)
• Timezone support
• SSE Streaming API with chunk_size/chunk_interval
• Async CSV/JSON download API
• Built-in anomaly detection (Z-Score, IQR, Moving Average, Auto)
• Built-in forecasting (SMA, Exponential, Linear, Holt-Winters, ARIMA, Prophet, Auto)
• Post-processing pipeline (downsampling: LTTB, MinMax, Average, M4)
• Group-scoped replication with startup sync and anti-entropy
• Node registration with etcd lease-based heartbeats
• Docker deployment
• Documentation site (Docusaurus)

In Progress

• Web-based admin dashboard
• Grafana plugin
• S3/GCS cold storage tier

Planned

• Prometheus remote write/read integration
• Automated backup and restore
• Multi-tenancy
• SQL-like query language

Contributing

We welcome contributions! Please see the GitHub repository for details.