Sentinel - Error Monitoring
Sentinel is error monitoring that focusing on simplicity, and holistic integration. Track backend services errors, let support raise issue ticket and link them to live service errors and let ai triage the errors
Error tracking
Debugging is complex, tiresome and quickly become nightmares if the issue is in flow that spans to different services. One of tiresome things in debugging are crawling the logs to find the gem (the error log). Existing solutions are diverse and feature complete, but for bootstrapped company / startups, but their free quota or paid plans quickly become infeasible. Of course, there are open source solutions, but based on our needs which requires high system efficiency (we don’t want our internal app become one of the cost center) we can’t use them
This project initiated from above problem. Since I am bootstrapping the service from scratch I can also add features that needed in my company flawlessly. Tech Support now can raise an issue in Sentinel and monitor their activities, got notified when there are updates. Engineer can link actual service errors with the issue that has been raised by Tech Support. Optionally AI also can help triaging, and links the issues automatically.
Tech Stack
- Backend: Golang, gRPC
- Frontend: Sveltekit
- AI: Google Gemini API
- Databases: PostgreSQL, Redis
- Queue: NATS
- Message Brokers: NATS Jetstream
- Infrastructure: Huawei Cloud, Docker, Kubernetes
Impact
- Reduced time to bug discovery for issues that raised by Tech Support
- Changed new bug discoveries from reactive to proactive because of automatic alerts from Sentinel
- Reduced time to debug recurring issues because issue linking and triaging