A Disastrous Day in the Life of a PostgreSQL DBA
A Disastrous Day in the Life of a PostgreSQL DBA

Presented by:

48287b84de29daed7d09d864dd9c2614

Kanthanathan S

from IITM Pravartak

Kanthan has over 32 years of experience in various practices ranging in building private clouds, Information Security, High performance parallel computing and complex web-scale application deployment architecture. Having worked across various clientlele in Defence, Banking and Financial Insitutions, Large Research Insitutions & Government projects, he has acquired knowledge in various nuances of plan, building and operational management of large scale systems.

4a086c8e303938170e0e31e74daf8047

Sathakkathullah

from IITM Pravartak

Sathak with over 30 years’ experience, possesses deep expertise in open-source databases, architecting high-performance relational systems for enterprises. With extensive experience as a technology leader in technology consulting and innovation, he has transformed prototypes into scalable products through strategic technology selection and resolution of challenges like legacy system modernization. Having engaged across multiple domains, he has honed skills in end-to-end technical leadership, user-centric design, and fostering open-source ecosystems to drive enterprise advancement, complemented by his education from the National University of Singapore.

No video of the event yet, sorry!

This 3-hour workshop simulates a high-pressure day of PostgreSQL outages—application slowdowns, storage spikes, WAL explosions, corruption alerts, and accidental data loss. Participants practice real-world troubleshooting using PostgreSQL logs, pg_stat_activity, lock diagnostics, autovacuum and bloat handling, WAL and checkpoint analysis, live query tuning, and backup/PITR-based recovery. The session also covers practical tasks like fixing stuck archiving, validating backups, resolving corruption at the relation level, terminating problem sessions, and performing safe reorganizations.

Learning Objectives:

  • Identify symptoms of common PostgreSQL outages
  • Map real-world incidents to PostgreSQL internals
  • Analyze PostgreSQL logs to detect root causes of performance and stability issues
  • Use emergency tools (pg_stat_activity, pg_locks, pg_repack, etc.) to mitigate live incidents
  • Handle storage crises and vacuum-related performance degradation
  • Learn from real-world war stories to build intuition and preventive habits

Date:
Duration:
3 h 30 min
Room:
Conference:
PGConf India, 2026
Language:
Track:
Training
Difficulty:
Medium