logo
Sign in

Observability

Gain comprehensive insights into your compute instance performance with detailed metrics, logs, and resource monitoring.

Performance Metrics

The dedicated job view allows correlating job steps with performance metrics and comparing a job's performance to previous runs. This visibility helps you identify performance bottlenecks and track improvements over time.

Instance metrics
  • CPU Usage: Track processor utilization across all cores, identify CPU-intensive operations
  • Memory Consumption: Detect memory leaks and excessive allocation
  • Disk I/O: Understand disk read/write patterns, identify storage bottlenecks
  • Network Activity: Track network usage, bandwidth consumption, and connection patterns

Advanced Logging Capabilities

Access comprehensive logging, providing visibility into every aspect of your workflow execution. Namespace log collection and serving goes beyond stdout/stderr and provides meaningful insights.

Container Logs Namespace provides direct access to logs from running containers within your workflows.

Container Logs

Build logs Any Docker build issued from a Namespace runner is linked from the instance UI, granting you direct access to the build logs and build performance telemetry.

Container Builds
Build Tracing

Out of Memory (OOM) Detection

Memory consumption graphs can already give insights on memory leaks and excessive allocation patterns. However, when only working with memory allocation metrics it is hard to isolate if a task was running close to a memory limit, or encountered an error when running out of memory. Namespace's automatic OOM detection makes these conditions explicit and simple to detect.

Crash Dump Collection

Namespace can automatically detect if your task execution crashes, collect crash dumps, and store them as artifacts. These dumps contain invaluable forensic information useful for determining the cause of the crash. Crash dump collection is not enabled by default - reach out to support@namespace.so to get access.

Hands-on Support

Got stuck? Need help with debugging one of your workflows? Our team is here to assist:

  • Technical Support: Reach out to support@namespace.so to talk to one of our engineers.
  • Community: Join our community Discord to learn about tips and best practices.