A collection of DTrace-based tools for real-time PostgreSQL performance monitoring without requiring configuration changes.
pgsql tools is a collection of command-line utilities that use PostgreSQL's built-in DTrace probes to monitor database performance in real-time. It provides detailed insights into query execution, transaction behavior, lock contention, and system metrics without requiring configuration changes or service restarts.
Database administrators and developers working with PostgreSQL in production environments who need immediate visibility into performance issues without disrupting service.
These tools offer zero-configuration monitoring that instruments all PostgreSQL processes on a system, providing detailed performance data that's difficult to obtain through standard logging or configuration-based monitoring approaches.
report top-level postgres stats
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Uses PostgreSQL's built-in DTrace probes, so no configuration changes or restarts are needed, as highlighted in the README's philosophy and tool descriptions.
Tools like pgsqlstat and pgsqlslower provide immediate metrics on queries, transactions, and locks at configurable intervals, ideal for live troubleshooting.
pgsqltxslower traces individual query phases (parse, plan, execute) and gaps between operations, offering granular visibility into transaction timelines.
Leverages native DTrace probes to instrument all PostgreSQL processes without affecting database operations, as emphasized in the README's implementation notes.
Requires DTrace support on the system and appropriate privileges, which the README warns can lead to no output if not configured, limiting portability.
The README explicitly states 'The output format for all of these tools may change,' indicating potential breaking changes and lack of backward compatibility.
Tools only monitor queries that start and finish during their runtime; very long-running queries may be missed, as noted in descriptions for pgsqlslower and pgsqllat.
Output from tools like pgsqltxslower can be out of order and require manual sorting, as mentioned in the caveats, adding to the analysis overhead.