In this chapter, we will learn several techniques to track sources of poor performance or stop potential outages before they occur. We will cover the following recipes in this chapter:
- Performing triage
- Installing common statistics packages
- Evaluating the current disk performance with iostat
- Tracking I/O-heavy processes with iotop
- Viewing past performance with sar
- Correlating performance with dstat
- Interpreting /proc/meminfo
- Examining /proc/net/bonding/bond0
- Checking the pg_stat_activity view
- Checking the pg_stat_statements view
- Deciphering database locks
- Debugging with strace
- Logging checkpoints properly
