Logoric security and architecture
How ingestion, keys, storage, project isolation and deployment options work for teams that need trust before connecting logs.
Before connecting logs, teams need to understand where data is stored, how keys work and which services are public or internal.
Logoric supports SaaS and self-hosted deployments, and GPU inference can be moved to a separate local HTTPS endpoint.
What is protected
- Project API keys are scoped for ingestion and can be disabled.
- Projects and spaces isolate member access and data.
- Demo incidents are marked as demo and do not consume user limits.
Architecture
- Frontend, backend, ingestion, Redis/Kafka/ClickHouse/Postgres run in the main Docker stack.
- GPU services vLLM and embed-api can run locally and be connected over HTTPS.
- Workers with database access should not be exposed publicly without private network or VPN controls.
Quick start
- Create a project automatically after signup.
- Generate a starter key and send the first curl.
- Validate live stream, alert and incident/RCA before production log rollout.
Security and architecture FAQ
Can the main server run without GPU?
Yes. The main product can run without GPU containers while vLLM/embed-api stay on a local GPU machine.
Do we need to send real production logs immediately?
No. Start with a test log or demo incident to validate the workflow safely.
Is self-hosted supported?
Yes. The architecture supports self-hosted and enterprise deployments where storage control and network restrictions matter.
Related pages
Validate Logoric safely with the first project
Start with a demo or test curl before connecting real production logs.