Advanced
PostgreSQL Internals Studio
Peek inside buffer tags, visibility maps, and background workers with guided traces. You will correlate pg_stat_* views with what you see on disk and practice safe observation techniques on staging stacks sized for SMB teams.
Duration: 6 weeks · 52 hours · Format: Hybrid workshops
Price (informational): ￥198,000
Request information
What is included
- Buffer pin tracing with documented stop conditions
- Checkpoint spikes: measurement and mitigation menu
- Autovacuum tuning without turning knobs blindly
- WAL generation exercises with controlled bursts
- Reading stack traces from backend logs
- Staging harness for repeatable load scripts
- Ethical load testing guardrails for shared clusters
Outcomes
- Produce a one-page health brief from stats views
- Recommend two vacuum-related changes with risk notes
- Document a checkpoint event with before/after graphs
Lead contact
Jonas Meyer
Database reliability consultant; contributes to internal Postgres training playbooks.
FAQ
We discuss sysctl at a high level only. Production changes stay with your platform team.
Most labs run on vanilla Postgres. Managed differences are called out in side notes.
We do not cover proprietary fork internals beyond compatibility footnotes.
Participant notes
The visibility map lab mirrored our staging incident. I still wish there was one more day on multixact edge cases.
Mika · Retail analytics vendor · 4/5 · Google
Checkpoint spike exercise used our anonymized metrics template — rare for a public syllabus.
Elena · Staff engineer