DOC-14534 Document possible disparity between P99 latency and Max latency (#20127)

* In statements-table.md and statement-details.md, updated P50, P90, P99 description with concise note about data source for latency percentiles.

* In statements-table.md and statement-details.md for v25.2 and v25.3, removed columns for P50, P90, P99.

Author: florence-crl

src/current/_includes/v23.1/ui/statements-table.md

@@ -14,9 +14,9 @@ Statement Time | Average [planning and execution time]({{ link_prefix }}architec
 <a id="percent-of-all-runtime"></a>% of All Runtime | The percentage of execution time taken by this statement fingerprint compared to all other statements executed within the time period, including those not displayed. Runtime is calculated as the mean execution latency multiplied by the execution count.<br><br><b>Note:</b> The sum of the values in this column may not equal 100%. Each fingerprint's percentage is calculated by dividing the fingerprint's runtime by the sum of the runtimes for all statement fingerprints in the time interval. "All statement fingerprints" means all user statement fingerprints (not only those displayed by the [search criteria](#search-criteria)), as well as internal statement fingerprints that are never included in the displayed result set. The search criteria are applied after the `% of All Runtime` calculation.
 Contention Time | Average time statements with this fingerprint were [in contention]({{ link_prefix }}performance-best-practices-overview.html#understanding-and-avoiding-transaction-contention) with other transactions within the time interval.<br><br>The gray bar indicates mean contention time. The blue bar indicates one standard deviation from the mean. Hover over the bar to display exact values.
 CPU Time | Average CPU time spent executing within the specified time interval. The gray bar indicates mean CPU time. The blue bar indicates one standard deviation from the mean. <br><br>The CPU time includes time spent in the [SQL layer]({{ link_prefix }}architecture/sql-layer.html). It does not include time spent in the [storage layer]({{ link_prefix }}architecture/storage-layer.html).
-P50 Latency | The 50th latency percentile for sampled statement executions with this fingerprint.
-P90 Latency | The 90th latency percentile for sampled statement executions with this fingerprint.
-P99 Latency | The 99th latency percentile for sampled statement executions with this fingerprint.
+P50 Latency | The 50th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P90 Latency | The 90th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P99 Latency | The 99th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
 Min Latency | The lowest latency value for all statement executions with this fingerprint.
 Max Latency | The highest latency value for all statement executions with this fingerprint.
 Rows Processed | Average number of rows read and written while executing statements with this fingerprint within the time interval.

src/current/_includes/v23.2/ui/statement-details.md

@@ -89,9 +89,9 @@ Average Rows Read | The average number of rows read when the plan was executed.
 Full Scan | Whether the execution performed a full scan of the table.
 Min Latency | The lowest latency value for all statement executions with this Explain Plan.
 Max Latency | The highest latency value for all statement executions with this Explain Plan.
-P50 Latency | The 50th latency percentile for sampled statement executions with this Explain Plan.
-P90 Latency | The 90th latency percentile for sampled statement executions with this Explain Plan.
-P99 Latency | The 99th latency percentile for sampled statement executions with this Explain Plan.
+P50 Latency | The 50th latency percentile for sampled statement executions with this Explain Plan.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P90 Latency | The 90th latency percentile for sampled statement executions with this Explain Plan.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P99 Latency | The 99th latency percentile for sampled statement executions with this Explain Plan.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
 Distributed | Whether the execution was distributed.
 Vectorized | Whether the execution used the [vectorized execution engine]({{ link_prefix }}vectorized-execution.html).
 

src/current/_includes/v23.2/ui/statements-table.md

@@ -14,9 +14,9 @@ Statement Time | Average [planning and execution time]({{ link_prefix }}architec
 <a id="percent-of-all-runtime"></a>% of All Runtime | The percentage of execution time taken by this statement fingerprint compared to all other statements executed within the time period, including those not displayed. Runtime is calculated as the mean execution latency multiplied by the execution count.<br><br><b>Note:</b> The sum of the values in this column may not equal 100%. Each fingerprint's percentage is calculated by dividing the fingerprint's runtime by the sum of the runtimes for all statement fingerprints in the time interval. "All statement fingerprints" means all user statement fingerprints (not only those displayed by the [search criteria](#search-criteria)), as well as internal statement fingerprints that are never included in the displayed result set. The search criteria are applied after the `% of All Runtime` calculation.
 Contention Time | Average time statements with this fingerprint were [in contention]({{ link_prefix }}performance-best-practices-overview.html#understanding-and-avoiding-transaction-contention) with other transactions within the time interval.<br><br>The gray bar indicates mean contention time. The blue bar indicates one standard deviation from the mean. Hover over the bar to display exact values.
 CPU Time | Average CPU time spent executing within the specified time interval. The gray bar indicates mean CPU time. The blue bar indicates one standard deviation from the mean. <br><br>The CPU time includes time spent in the [SQL layer]({{ link_prefix }}architecture/sql-layer.html). It does not include time spent in the [storage layer]({{ link_prefix }}architecture/storage-layer.html).
-P50 Latency | The 50th latency percentile for sampled statement executions with this fingerprint.
-P90 Latency | The 90th latency percentile for sampled statement executions with this fingerprint.
-P99 Latency | The 99th latency percentile for sampled statement executions with this fingerprint.
+P50 Latency | The 50th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P90 Latency | The 90th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P99 Latency | The 99th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
 Min Latency | The lowest latency value for all statement executions with this fingerprint.
 Max Latency | The highest latency value for all statement executions with this fingerprint.
 Rows Processed | Average number of rows read and written while executing statements with this fingerprint within the time interval.

src/current/_includes/v24.1/ui/statement-details.md

@@ -89,9 +89,9 @@ Average Rows Read | The average number of rows read when the plan was executed.
 Full Scan | Whether the execution performed a full scan of the table.
 Min Latency | The lowest latency value for all statement executions with this Explain Plan.
 Max Latency | The highest latency value for all statement executions with this Explain Plan.
-P50 Latency | The 50th latency percentile for sampled statement executions with this Explain Plan.
-P90 Latency | The 90th latency percentile for sampled statement executions with this Explain Plan.
-P99 Latency | The 99th latency percentile for sampled statement executions with this Explain Plan.
+P50 Latency | The 50th latency percentile for sampled statement executions with this Explain Plan.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P90 Latency | The 90th latency percentile for sampled statement executions with this Explain Plan.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P99 Latency | The 99th latency percentile for sampled statement executions with this Explain Plan.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
 Distributed | Whether the execution was distributed.
 Vectorized | Whether the execution used the [vectorized execution engine]({{ link_prefix }}vectorized-execution.html).
 

src/current/_includes/v24.1/ui/statements-table.md

@@ -14,9 +14,9 @@ Statement Time | Average [planning and execution time]({{ link_prefix }}architec
 <a id="percent-of-all-runtime"></a>% of All Runtime | The percentage of execution time taken by this statement fingerprint compared to all other statements executed within the time period, including those not displayed. Runtime is calculated as the mean execution latency multiplied by the execution count.<br><br><b>Note:</b> The sum of the values in this column may not equal 100%. Each fingerprint's percentage is calculated by dividing the fingerprint's runtime by the sum of the runtimes for all statement fingerprints in the time interval. "All statement fingerprints" means all user statement fingerprints (not only those displayed by the [search criteria](#search-criteria)), as well as internal statement fingerprints that are never included in the displayed result set. The search criteria are applied after the `% of All Runtime` calculation.
 Contention Time | Average time statements with this fingerprint were [in contention]({{ link_prefix }}performance-best-practices-overview.html#understanding-and-avoiding-transaction-contention) with other transactions within the time interval.<br><br>The gray bar indicates mean contention time. The blue bar indicates one standard deviation from the mean. Hover over the bar to display exact values.
 SQL CPU Time | Average SQL CPU time spent executing within the specified time interval. It does not include SQL planning time or KV execution time. The gray bar indicates mean SQL CPU time. The blue bar indicates one standard deviation from the mean. <br><br>The SQL CPU time includes time spent in the [SQL layer]({{ link_prefix }}architecture/sql-layer.html). It does not include time spent in the [storage layer]({{ link_prefix }}architecture/storage-layer.html).
-P50 Latency | The 50th latency percentile for sampled statement executions with this fingerprint.
-P90 Latency | The 90th latency percentile for sampled statement executions with this fingerprint.
-P99 Latency | The 99th latency percentile for sampled statement executions with this fingerprint.
+P50 Latency | The 50th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P90 Latency | The 90th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P99 Latency | The 99th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
 Min Latency | The lowest latency value for all statement executions with this fingerprint.
 Max Latency | The highest latency value for all statement executions with this fingerprint.
 Rows Processed | Average number of rows read and written while executing statements with this fingerprint within the time interval.

src/current/_includes/v24.2/ui/statement-details.md

@@ -89,9 +89,9 @@ Average Rows Read | The average number of rows read when the plan was executed.
 Full Scan | Whether the execution performed a full scan of the table.
 Min Latency | The lowest latency value for all statement executions with this Explain Plan.
 Max Latency | The highest latency value for all statement executions with this Explain Plan.
-P50 Latency | The 50th latency percentile for sampled statement executions with this Explain Plan.
-P90 Latency | The 90th latency percentile for sampled statement executions with this Explain Plan.
-P99 Latency | The 99th latency percentile for sampled statement executions with this Explain Plan.
+P50 Latency | The 50th latency percentile for sampled statement executions with this Explain Plan.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P90 Latency | The 90th latency percentile for sampled statement executions with this Explain Plan.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P99 Latency | The 99th latency percentile for sampled statement executions with this Explain Plan.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
 Distributed | Whether the execution was distributed.
 Vectorized | Whether the execution used the [vectorized execution engine]({{ link_prefix }}vectorized-execution.html).
 

src/current/_includes/v24.2/ui/statements-table.md

@@ -14,9 +14,9 @@ Statement Time | Average [planning and execution time]({{ link_prefix }}architec
 <a id="percent-of-all-runtime"></a>% of All Runtime | The percentage of execution time taken by this statement fingerprint compared to all other statements executed within the time period, including those not displayed. Runtime is calculated as the mean execution latency multiplied by the execution count.<br><br><b>Note:</b> The sum of the values in this column may not equal 100%. Each fingerprint's percentage is calculated by dividing the fingerprint's runtime by the sum of the runtimes for all statement fingerprints in the time interval. "All statement fingerprints" means all user statement fingerprints (not only those displayed by the [search criteria](#search-criteria)), as well as internal statement fingerprints that are never included in the displayed result set. The search criteria are applied after the `% of All Runtime` calculation.
 Contention Time | Average time statements with this fingerprint were [in contention]({{ link_prefix }}performance-best-practices-overview.html#understanding-and-avoiding-transaction-contention) with other transactions within the time interval.<br><br>The gray bar indicates mean contention time. The blue bar indicates one standard deviation from the mean. Hover over the bar to display exact values.
 SQL CPU Time | Average SQL CPU time spent executing within the specified time interval. It does not include SQL planning time or KV execution time. The gray bar indicates mean SQL CPU time. The blue bar indicates one standard deviation from the mean. <br><br>The SQL CPU time includes time spent in the [SQL layer]({{ link_prefix }}architecture/sql-layer.html). It does not include time spent in the [storage layer]({{ link_prefix }}architecture/storage-layer.html).
-P50 Latency | The 50th latency percentile for sampled statement executions with this fingerprint.
-P90 Latency | The 90th latency percentile for sampled statement executions with this fingerprint.
-P99 Latency | The 99th latency percentile for sampled statement executions with this fingerprint.
+P50 Latency | The 50th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P90 Latency | The 90th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
+P99 Latency | The 99th latency percentile for sampled statement executions with this fingerprint.<br><br><strong>Note:</strong> Latency percentiles are calculated from a different set of executions than other statistics, so they may be inconsistent with the minimum, maximum, and average values. This data is provided for reference only.
 Min Latency | The lowest latency value for all statement executions with this fingerprint.
 Max Latency | The highest latency value for all statement executions with this fingerprint.
 Rows Processed | Average number of rows read and written while executing statements with this fingerprint within the time interval.