Plan Explorer Index Analysis

The Plan Explorer Index Analysis tab provides a sandbox environment for planning indexing strategy relative to a specific operation in a query.

For Index Analysis to populate, the query must execute from within the Plan Explorer session, using the Get Actual Plan menu option.

The Statement Tree and Plan Diagram

For more information, see the Integrated Plan Explorer topic. 

The Text Data Area

For more information, see the Integrated Plan Explorer topic. 

The Indexes Data Grid

SentryOne Indexes Data Grid

The top row is the index score that's based on several factors including covering, value density, sorting effectiveness, and seekability. 100 percent is the best possible score; however, there are some situations that prevent the possibility of a 100 percent score. One known cause is when columns are missing statistics. 

Column statistics are provided for each table column used by the query, with table columns used by the selected operation displayed in bold.

Data Grid Columns:

SentryOne Data Grid Columns

NameDescription
Table ColumnName of the column in the relevant database table.
DensityHow selective the row values are for the column (lower values are more selective).
Last Statistics UpdateThe date the column statistics were last updated.
Update StatisticsAutomatically updates the existing statistics. 
Avg. LengthAverage length of data values.
Estimated Size (MB)The estimated size of the column values (in MB).
OutputChecked if the column is an output for the selected operation.
Sort TypeSort order (ASC, DESC).
Sort OrderOrdinal position of the column in the sort.
PredicateThe search or join predicate(s) applied to the table column by the operation.

Columns in the grid to the right of the Predicate column all represent indexes. These include both special case indexes, as well as the indexes defined on the table.

Indexes can be modified, as well as scripted for create, drop, or drop and create. Manually update statistics using the button at the bottom of the index column.

Index Coloring:

  • Shades of green indicate that the ordinal position of the column is such that a favorable operation may occur, such as a seek over a scan. Darker shades of green indicate that the column is filtered.
  • Yellow indicates that the column is used in the query, and covered, but isn't part of the left subset of the index key. Included columns are also listed in yellow.
  • Red indicates that the column involved isn't covered by the index.

The Parameters Data Grid

The parameters grid lists parameters used by the query. This differs from the Plan Explorer Parameters tab in that it uses tipping point logic to help determine the type of operation (Seek or Scan) that the optimizer might choose for that parameter value. This isn't an exact science, and should be used as a rough guideline.

The parameters grid provides a Test Value column that allows you to enter a new parameter value and asks for a new Estimated Plan. This allows you to test different parameter values against the plan to see if the plan shape changes given that parameter value.

The small arrows to the right of the parameter values allow you to quickly place the value of that parameter into the Test Values column.

SentryOne Parameters Data Grid

The Statistics Histogram Chart

The Histogram chart displays the histogram steps for the index statistics visually. This allows you to quickly spot potential parameter sniffing issues.

This chart also works in conjunction with the parameters grid, and selected parameters that are present in the histogram have their histogram buckets highlighted on the chart. This also works for parameter ranges.

SentryOne Histogram Chart

Note:  For more information about Index Analysis, see the Sentry v10 : Index Analysis blog post.