Beyond Alerts: Using PostgreSQL for Operational Log Analysis
Presented by:
Diksha Sharma
Diksha Sharma is a technology professional at Fujitsu Australia Limited with a Master of Data Science, specialising in enterprise PostgreSQL support and practical AI initiatives. Her experience includes diagnosing complex technical issues, supporting customer environments, and developing data-driven solutions that combine database technologies with emerging AI capabilities.
No video of the event yet, sorry!
Monitoring tools are effective at detecting when a server or PostgreSQL service is unhealthy, but they do not always explain why an incident occurred. Engineers may still need to search through large log files, compare timestamps, and connect related events across multiple sources before the cause becomes clear.
This session explores how PostgreSQL can be used as an investigation platform for operational logs. Using a PostgreSQL service failure on a Linux server as a practical case study, it demonstrates how raw OS log messages can be ingested into PostgreSQL, transformed into structured events, and analysed using SQL, full-text search, and vector similarity with pgvector.
The session follows the investigation from the initial alert and original log messages through to a structured, searchable incident timeline presented in a simple investigation interface. It shows how exact keyword searches, metadata filters, and semantic similarity can work together to uncover related events that may use different terminology.
The aim is not to replace existing monitoring or observability platforms, but to demonstrate how PostgreSQL can complement them by making selected operational data easier to query, correlate, and understand.
Designed for PostgreSQL DBAs, developers, support engineers, and platform teams, this session provides a practical and repeatable approach for turning unstructured logs into structured incident evidence and supporting deeper technical investigation.
- Date:
- Duration:
- 40 min
- Room:
- Conference:
- PG Down Under 2026
- Language:
- Track:
- Ops and Administration
- Difficulty:
- Medium