EDB Stat Monitor 2.1.0 release notes

Released: 12 June 2025

This is the First public release of EDB Stat Monitor (edb_stat_monitor), a new statistics monitoring tool. It's built on the foundation of pg_stat_monitor. edb_stat_monitor introduces enhanced capabilities for query performance analysis.

Highlights

  • New sampling logic with setting for edbsm_sample_rate.
  • More precise calculation of quantiles.
  • Controllable compression levels.
  • New functions: edb_stat_monitor_td_histogram and edb_stat_monitor_td_quantile.

Features

DescriptionAddresses
First public release of EDB Stat Monitor (edb_stat_monitor)
Added digest logic

The digest logic has been added to the collection of execution time data, allowing for the calculation of more precise quantiles. This enhancement improves the accuracy of query performance analysis by providing better insights into execution time distribution.

Added sampling logic

Captures a subset of incoming queries. This provides a representative "taste" of system activity, offering crucial insights into performance and usage.

Added GUC parameter edbsm_td_compression

A new GUC parameter, edbsm_td_compression, has been introduced to control the compression level of the t-digest data structure. This allows users to optimize the balance between memory usage and accuracy in execution time quantile calculations.

Added GUC parameter edbsm_sample_rate

The new GUC parameter edbsm_sample_rate has been added to control the sampling rate for query execution statistics. This feature allows users to adjust the frequency of data collection, enabling more efficient resource usage while still capturing essential performance metrics.

New function edb_stat_monitor_td_histogram

The new function edb_stat_monitor_td_histogram has been added to display the t-digest histogram. This function provides a visual representation of the execution time distribution, helping users to better understand query performance patterns.

New function edb_stat_monitor_td_quantile

The new function edb_stat_monitor_td_quantile has been introduced to calculate percentiles using the t-digest algorithm. This function allows users to obtain precise quantile values for execution times, enhancing performance analysis capabilities.


Could this page be better? Report a problem or suggest an addition!