Systems exhibition01 / Problem 02 / Architecture 03 / How it works 04 / Failure thinking 05 / Reflection Inspect the sourceLog Intelligence Tool on GitHub
Diagnostic workflow · Structured analysis
Log Intelligence Tool
A rule-driven analysis service that correlates database traces, kernel errors, and Java stack traces into structured root-cause reports.Java 17Spring BootPostgreSQLRESTDockerPrometheusGrafanaOpenTelemetry
ORA-00600 → KERNELdeadlock → CONCURRENCYGC overhead → MEMORYlogsparsepatternsdiagnostics
The constraint.
Production incident triage means correlating signals that live in different places — application logs, database traces, kernel errors — and doing it manually under time pressure.
How it is shaped.
A Spring MVC API feeds type detection, specialized parsers, pattern matching, layer classification, and root-cause analysis. Uploaded files run on a bounded async executor; PostgreSQL stores history.
Decisions made explicit.
- Format-specific parsers behind one pipeline
- Async file processing with pollable status
- Deterministic patterns and explainable remediation text
Where it can break.
Uploads begin as PENDING and transition to COMPLETED or FAILED. Validation and structured API errors keep bad inputs out of the parser pipeline.
What I learned.
Diagnostic automation is credible when every conclusion can be traced back to evidence in the input.