How to deal with long-running transactions (OLTP)
Content from #0030 to be migrated from GitLab repository
Long-running transactions are the "heart attack" scenario in OLTP systems. This guide provides emergency response procedures and prevention strategies for transactions that can bring your production system to its knees.
What you'll learn​
- How to identify long-running transactions
- Emergency procedures to kill problematic transactions
- Understanding transaction impact on system performance
- Prevention strategies and monitoring
- Recovery procedures after transaction termination
Prerequisites​
- Access to pg_stat_activity
- Understanding of transaction concepts
- Ability to terminate database connections
This content will be migrated from the original GitLab repository guide #0030.