// course
Data Pipelines for Logs
Buffering, parsing, and cardinality control for noisy syslog streams.
7,200,000 VND (informational)
Overview
Ingest synthetic bursts, tame rogue fields, and build a dashboard that does not melt GPUs. Mentors grade on clarity, not chart flashiness.
Included focus areas
- Docker-compose lab stack
- Parser tests with golden files
- Sampling vs drop policies exercise
- Retention tier spreadsheet template
- SLO draft for search latency
- On-call runbook snippet
- Cost estimate worksheet
Outcomes
- Cut indexed field count on a sample stream
- Write a runbook entry for parser failure
- Present a sane retention proposal
Lead mentor
Nhat Linh Dang
Observability engineer; hates expensive dashboards.
FAQ
We teach patterns; week one picks a stack for exercises based on cohort vote.
Experience notes
Cardinality lab embarrassed our old “log everything” default — politely.