Overview

The Metrics view helps you understand both the current health of an Iceberg table and the impact of optimization runs.

There are two types of metrics:

• Table Metrics – always available for a table, showing its current state.
• Run Metrics – measurements for a single optimization run: input vs output data and delete files (counts and sizes) for that run.

Table Metrics

You can see table-level metrics from the Tables page by clicking the View Metrics button for a table.

These metrics describe the current layout and size of the Iceberg table:

Metric	Description
File Count	Total number of data and delete files that currently make up the table.
Total Size	Combined on-disk size of all files that belong to the table.
Average File Size	Average size of files; useful to see if the table has many small files.
Last Commit Time	Time of the last successful Iceberg commit for the table.
Data Files Count	Number of data files that store actual table rows.
Delete Files Count	Number of delete files(equality+position) that track removed or filtered-out rows.
Health Score	Combined score based on Small Files Score, Eq Delete Score, and Pos Delete Score.

Health Score appears in the Tables page, not in the View Metrics modal.

Run Metrics

Run Metrics describe what one optimization run did to the table: how many data and delete files went in, and how many data and delete files came out. Open them from the Runs page via the View Metrics action for the run you want to inspect.

These metrics help you see how much a specific run improved the table:

Size units

All size fields in Table metrics and Run metrics are shown in bytes.

Metric	Description
Input data files	Count of input data files
Input data size	Total size of input data files
Input equality delete files	Count of input equality delete files (Count of equality delete files in the current snapshot * No.of sub-tasks)
Input equality delete size	Total size of input equality delete files
Input position delete files	Count of input position delete files (Count of position delete files in the current snapshot * No.of sub-tasks)
Input position delete size	Total size of input position delete files
Output data files	Count of output data files produced
Output data size	Total size of output data files produced
Output delete files	Count of output delete files produced
Output delete size	Total size of output delete files produced

Interpreting Metrics

• A high Delete Files Count or very low Average File Size can indicate fragmentation and metadata overhead.
• After a successful optimization run, compare Input vs Output.
• A smaller Output delete size (and/or fewer Output delete files) typically indicates the run reduced delete-file fragmentation.
• Changes in Input/Output data file counts and data size show how much data was rewritten/compacted for that run’s scope.