Disk Usage

functional

Size every database on the instance, plus the current database's file layout.

Opened from a server, Disk Usage gives you the whole instance: a Databases tab that sizes every database - largest first, plus, on SQL Server and ClickHouse, a Files (current database) tab for the storage behind it. The columns follow what each engine can actually report.

Opening it

  • Right-click a server in the explorer and choose Tools › Disk Usage....
  • Run Disk Usage… from the command palette.
  • Use the Disk Usage link in the Server Dashboard's footer.

The pane runs its queries on the connection you opened it from. Refresh re-reads the sizes, and Pop out ⤢ in the title bar moves the pane into its own window so you can keep it beside your work. A line under the grid counts the rows returned.

What each engine reports

EngineDatabases tab
SQL Server, Azure SQLName, state, recovery model, data MB, log MB, and file count per database, sorted by data size.
PostgreSQLName, owner, and size in MB. Template databases are left out, and databases your login cannot connect to show an honest blank size instead of a guess.
MySQL, MariaDBData, index, and free MB per schema, with a table count. Built-in schemas such as mysql are included.
SQLite, libSQL / TursoThe file itself: size, reclaimable space (free pages), page count, and page size, from the engine's page counters.
DuckDBOne row per attached database from pragma_database_size(): size in use, space a CHECKPOINT could reclaim, write-ahead log size, block counts and size, and memory in use against the memory limit.
ClickHouseEvery database sized from its active parts in system.parts: bytes on disk, the uncompressed size those bytes hold, rows, tables, and part count.
OracleThe storage your login's schema owns, per tablespace, from USER_SEGMENTS: total, table, index, and LOB MB plus a segment count. It needs no DBA privilege, and it is not an instance-wide total.
RedisOne row per logical database: keys, keys with an expiry, and average TTL from INFO keyspace, plus memory measured with MEMORY USAGE on a SCAN sample of up to 200 keys, scaled to an estimate for the whole database, with the largest sampled key.

The Files tab

On SQL Server, the Files tab lists the current database's files: logical name, type (data or log), physical path, size, maximum size (blank when the file can grow without limit), and the growth increment as a percentage or in MB. On ClickHouse, the server's configured disks stand in for files: name, type, path, space used, and total space. Other engines have no Files tab.

Scope-aware. This is the server-wide view. Open Disk Usage from a single database instead and it drops the all-databases tab and shows only that database's files. See Disk Usage (database).
SQLly
SQLly server-scoped Disk Usage listing all databases
Server Disk Usage: the Databases tab sizing every database on the instance.

Reading the numbers

  • A big log next to a small database on SQL Server often means the log has not been backed up or truncated; the recovery model column tells you whether log backups are expected.
  • Reclaimable space on SQLite and DuckDB is space the file already holds but no longer uses. It stays in the file until you compact it - VACUUM on SQLite, CHECKPOINT on DuckDB.
  • Redis figures marked estimated come from the sample, so treat them as a guide. For memory by data type and the biggest keys, open the Memory analyzer from the database's right-click menu.
  • An error instead of a grid is the engine's own message, shown as-is, for example, a login without permission to read the catalog view the query uses.

Read-only, this pane only observes sizes and growth settings; it never changes them. Engines without a disk-usage query say Disk usage is not available on … rather than showing an empty grid.