deployment practice · Self-paced + demo Fridays
Observability Stack for JVM Services
OpenTelemetry wiring, exemplar traces, and RED metrics without drowning dashboards.
- Duration
- 4 weeks · 42 mentor hours
- Schedule
- Flexible
- Level
- Intermediate
- Price
- 11,800 THB (informational)
OpenTelemetry · Prometheus · Grafana
Description
You attach OTel auto-instrumentation, craft span attributes meaningful to product owners, and wire exemplars into Prometheus + Grafana stacks running locally. Labs emphasize tail sampling strategies.
What arrives in your repo
- OTel Java agent vs manual instrumentation matrix
- Baggage propagation exercises
- Grafana RED panel templates
- Log correlation with trace IDs
- SLO burn-rate alert examples
- Cost guardrails for high-cardinality labels
- Bangkok business-hour alert tuning
Outcomes we expect you to evidence
- Publish a service dashboard with SLOs
- Document cardinality budget
- Run a tabletop incident using traces
Responsible instructor
Pimchanok Lert
SRE coach focused on JVM noise reduction.
FAQ
Local only; bring your cloud if you want to replay labs.
We discuss migration from legacy agents but do not configure vendor APM.
No mobile RUM — server JVM only.
Experience notes
“Observability Stack for JVM Services cardinality budget worksheet stopped our label explosion.”