Exploring a Specific Table

The Table page shows the compute consumption and the number of queries run for a table during the reporting period, and includes detailed information about the table columns.

At the top of the Tables page, the graphs show the behavior of the top five workloads (by compute consumption and by number of queries) that hit the table. All the remaining workloads are aggregated in a sixth workload named Other.

In the CPU% graph:

  • The X-axis shows time.
  • The Y-axis shows the compute utilization for scanning the table relative to the total compute utilization for scanning all the tables.

In the Queries graph:

  • The X-axis shows time.
  • The Y-axis shows the number of queries executed on the table relative to the total number of queries.

Below the graph, the following information is listed for each column:

  • Column: The column name in the table.
  • Type: The underlying column type in the table, such as double, integer, or varchar.
  • Instructions: The existing acceleration instruction types defined for the column, either Data, Index, Lucene, Bloom or Not Accelerated. Multiple instruction types can be defined for a single column.
  • Popularity: The percentage of queries associated with the column relative to the total number of queries executed against the table.
  • Partitions: The number of partitions used in queries.

The columns are listed in descending order of the Popularity column. Click any of the column titles in the table to sort the columns, or type a string in the Search field to find a column by name.

👍

Select one or more columns, then click Accelerate to create or override the instruction‍ for the column(s).