26.12.17.4 Statement Histogram Summary Tables

The Performance Schema maintains statement event summary tables that contain information about minimum, maximum, and average statement latency (see Section 26.12.17.3, “Statement Summary Tables”). Those tables permit high-level assessment of system performance. To permit assessment at a more fine-grained level, the Performance Schema also collects histogram data for statement latencies. These histograms provide additional insight into latency distributions.

Section 26.12.6, “Performance Schema Statement Event Tables” describes the events on which statement summaries are based. See that discussion for information about the content of statement events, the current and historical statement event tables, and how to control statement event collection, which is partially disabled by default.

Example statement histogram information:

For example, in row 3, these values indicate that 23.52% of queries run in under 75.86 microseconds:

In row 4, these values indicate that 62.50% of queries run in under 79.44 microseconds:

Each statement histogram summary table has one or more grouping columns to indicate how the table aggregates events:

A histogram consists of N buckets, where each row represents one bucket, with the bucket number indicated by the BUCKET_NUMBER column. Bucket numbers begin with 0.

Each statement histogram summary table has these summary columns containing aggregated values:

The statement histogram summary tables have these indexes:

TRUNCATE TABLE is permitted for statement histogram summary tables. Truncation sets the COUNT_BUCKET and COUNT_BUCKET_AND_LOWER columns to 0.

In addition, truncating events_statements_summary_by_digest implicitly truncates events_statements_histogram_by_digest, and truncating events_statements_summary_global_by_event_name implicitly truncates events_statements_histogram_global.