Third-party cyber evaluations involving OpenAI models
openai.com·13h ago
Cloud & Infrastructure
Google Cloud Blog·Jun 23·5 min read
Log Analytics is now Observability Analytics: Query logs and traces with SQL
TL;DR
Developers and site reliability engineers (SREs) often struggle to gain comprehensive insights into application behavior due to fragmented logging and tracing tools. Google Cloud has introduced Observability Analytics, which integrates trace data and logs, allowing users to query them using SQL.
✦ Why It Matters
Engineers can leverage SQL queries to gain actionable insights from combined logs and traces, improving troubleshooting efficiency.
Key Takeaways
- Utilize SQL in Cloud Trace for advanced log and trace analysis.
- Identify performance issues in microservices with unified data.
- Analyze AI agent telemetry to improve tool reliability.
- Leverage the Observability API for programmatic workload management.
Full Summary
Effective application operation and troubleshooting require a clear understanding of system behavior, which is often hindered by disjointed logging and tracing tools. Google Cloud has launched Observability Analytics, which combines trace data and logs into a single platform, now generally available.
This suite includes the Observability API for management and configuration, allowing for programmatic control of workloads. A key feature is the support for SQL in Cloud Trace, enabling users to write complex queries that join application logs with distributed trace spans.
For example, users can identify checkout requests exceeding five seconds and determine which microservice is responsible for delays. Additionally, AI researchers can analyze telemetry data to find frequently failing tool calls or calculate performance metrics like the 95th percentile response time.
This integration enhances the ability to perform root-cause analysis and optimize application performance.