{"id":34819211,"url":"https://github.com/aqstack/sentinel","last_synced_at":"2026-01-13T20:47:42.857Z","repository":{"id":330427033,"uuid":"1122600499","full_name":"aqstack/sentinel","owner":"aqstack","description":"Self-healing edge computing agent with predictive failure detection and partition-resilient orchestration for Kubernetes","archived":false,"fork":false,"pushed_at":"2025-12-25T11:18:39.000Z","size":73,"stargazers_count":2,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-12-26T23:41:38.413Z","etag":null,"topics":["aiops","distributed-systems","edge-computing","failure-prediction","golang","kubernetes","machine-learning","prometheus","raft","self-healing"],"latest_commit_sha":null,"homepage":"","language":"Go","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/aqstack.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-12-25T05:22:51.000Z","updated_at":"2025-12-25T11:24:37.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/aqstack/sentinel","commit_stats":null,"previous_names":["aqstack/sentinel"],"tags_count":1,"template":false,"template_full_name":null,"purl":"pkg:github/aqstack/sentinel","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aqstack%2Fsentinel","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aqstack%2Fsentinel/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aqstack%2Fsentinel/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aqstack%2Fsentinel/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/aqstack","download_url":"https://codeload.github.com/aqstack/sentinel/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aqstack%2Fsentinel/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":28400125,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-01-13T14:36:09.778Z","status":"ssl_error","status_checked_at":"2026-01-13T14:35:19.697Z","response_time":56,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["aiops","distributed-systems","edge-computing","failure-prediction","golang","kubernetes","machine-learning","prometheus","raft","self-healing"],"created_at":"2025-12-25T14:03:56.858Z","updated_at":"2026-01-13T20:47:42.852Z","avatar_url":"https://github.com/aqstack.png","language":"Go","funding_links":[],"categories":["AI for *Ops","AIOps Platforms \u0026 Frameworks"],"sub_categories":["Automation \u0026 Self-Healing","Predictive Analytics for Operations"],"readme":"# Sentinel\n\n[![CI](https://github.com/aqstack/sentinel/actions/workflows/ci.yml/badge.svg?branch=main)](https://github.com/aqstack/sentinel/actions/workflows/ci.yml)\n[![Go Report Card](https://goreportcard.com/badge/github.com/aqstack/sentinel)](https://goreportcard.com/report/github.com/aqstack/sentinel)\n\nPredictive failure detection and autonomous orchestration for Kubernetes edge nodes.\n\n## What it does\n\nSentinel runs on each edge node and does three things:\n\n1. **Predicts failures** - Monitors CPU thermals, memory pressure, disk I/O, and network health. Uses lightweight statistical models to predict failures before they happen.\n\n2. **Survives partitions** - When nodes lose contact with the control plane, they form a local consensus group and continue operating autonomously.\n\n3. **Takes action** - Triggers preemptive workload migrations, cordons unhealthy nodes, and logs decisions for reconciliation when connectivity returns.\n\n## Why\n\nKubernetes assumes the control plane is always reachable. Edge environments break this assumption constantly - spotty networks, power issues, harsh conditions. When a node goes \"Unknown\", Kubernetes stops making decisions for it.\n\nCloud AIOps tools assume unlimited resources. Edge nodes run K8s + workloads + monitoring in 4GB RAM with no datacenter cooling.\n\nSentinel fills the gap.\n\n## What's different\n\nMost edge/IoT platforms focus on deployment and connectivity. Most AIOps tools assume cloud-scale resources. Sentinel combines ideas from both:\n\n- **Predictive, not reactive** - Catches thermal throttling and memory pressure before they cause failures\n- **Autonomous, not dependent** - Continues operating during control plane partitions instead of going dark\n- **Lightweight, not bloated** - Runs statistical models in \u003c64MB RAM, no GPUs or cloud inference needed\n- **Consensus-aware** - Nodes coordinate decisions locally using Raft-lite when disconnected\n\n## Quick Start\n\n```bash\n# Build\nmake build\n\n# Run locally\n./bin/predictor --node=my-node\n\n# With config file\n./bin/predictor --config=config.yaml\n\n# Deploy to cluster\nhelm install sentinel ./deploy/helm/sentinel\n```\n\n## Configuration\n\nSentinel supports YAML/JSON config files with CLI flag overrides:\n\n```yaml\nnode:\n  name: edge-node-1\n\nserver:\n  listen: \":9101\"\n  metrics: \":9100\"\n\npredictor:\n  interval: 1s\n  warn_threshold: 0.3\n  critical_threshold: 0.7\n  risk_weights:\n    thermal: 0.3\n    memory: 0.3\n    disk: 0.2\n    network: 0.2\n\nconsensus:\n  enabled: true\n  addr: \":9200\"\n  peers:\n    - edge-node-2:9200\n    - edge-node-3:9200\n\nlogging:\n  level: info\n  format: json\n```\n\nCLI flags:\n\n```bash\n--config          Config file path\n--node            Node name\n--listen          API listen address (default :9101)\n--metrics         Prometheus metrics address (default :9100)\n--interval        Collection interval (default 1s)\n--log-level       Log level: debug, info, warn, error\n--log-format      Log format: text, json\n--consensus-addr  Consensus listen address\n--peers           Comma-separated peer addresses\n```\n\n## Architecture\n\n```\n┌──────────────────────────────────────────────────────────┐\n│                        Sentinel                          │\n│                                                          │\n│  ┌────────────┐  ┌────────────┐  ┌───────────────────┐   │\n│  │ Collector  │─▶│ Predictor  │─▶│ Consensus (Raft)  │   │\n│  │            │  │            │  │                   │   │\n│  │ /proc, /sys│  │ ML model   │  │ Leader election   │   │\n│  │ thermals   │  │ risk calc  │  │ Decision log      │   │\n│  └────────────┘  └────────────┘  └───────────────────┘   │\n│        │               │                  │              │\n│        └───────────────┼──────────────────┘              │\n│                        ▼                                 │\n│            ┌─────────────────────┐                       │\n│            │ Prometheus Metrics  │                       │\n│            │ Health Endpoints    │                       │\n│            │ K8s Client          │                       │\n│            └─────────────────────┘                       │\n└──────────────────────────────────────────────────────────┘\n```\n\n## API\n\n| Endpoint | Description |\n|----------|-------------|\n| `/health` | Detailed health status with component checks |\n| `/healthz` | Liveness probe |\n| `/readyz` | Readiness probe |\n| `/prediction` | Current failure prediction |\n| `/metrics/latest` | Raw metrics snapshot |\n| `/consensus` | Consensus state and peers |\n| `/decisions` | Autonomous decisions log |\n| `/metrics` | Prometheus metrics |\n\n## Metrics\n\nPrediction:\n- `sentinel_prediction_failure_probability` - 0 to 1\n- `sentinel_prediction_confidence` - model confidence\n- `sentinel_prediction_time_to_failure_seconds` - estimated TTF\n- `sentinel_prediction_preemptive_migrations_total` - migration count\n\nPartition:\n- `sentinel_partition_detected` - 0 or 1\n- `sentinel_partition_duration_seconds` - how long partitioned\n- `sentinel_consensus_is_leader` - leader status\n- `sentinel_partition_autonomous_decisions_total` - decisions made offline\n\nRate limiting:\n- `sentinel_ratelimit_dropped_total` - messages dropped by rate limiter\n- `sentinel_ratelimit_enabled` - rate limiter state\n\n## Key Features\n\n**Circuit Breaker** - Protects against cascading failures when the K8s API is unreachable. Configurable thresholds with automatic recovery.\n\n**Exponential Backoff** - Peer reconnection uses backoff with jitter to prevent thundering herd during network recovery.\n\n**Rate Limiting** - Token bucket rate limiter on consensus messages to handle bursty traffic.\n\n**Structured Logging** - JSON or text output with log levels, component tags, and context propagation.\n\n**Health Checks** - Kubernetes-native liveness and readiness probes with detailed component status.\n\n## Development\n\n```bash\nmake test          # Run tests\nmake build         # Build for current platform\nmake build-arm64   # Build for ARM64 edge devices\nmake docker-build  # Build container image\n```\n\nRequires Go 1.21+.\n\n## License\n\nMIT\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faqstack%2Fsentinel","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Faqstack%2Fsentinel","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faqstack%2Fsentinel/lists"}