Skip to content
ClickHouse Docs
ClickHouse DocsClickHouse Docs
49.6kGet Started

system.keeper_snapshots

View as MarkdownEdit this pageReport an issueSet up docs MCP

Description

This table does not exist if this node is not configured to run an in-process ClickHouse Keeper. It contains one row per Raft snapshot file tracked by the in-process Keeper state machine, including snapshots currently being received from the leader.

Columns

  • last_log_index (UInt64) — Last log index covered by the snapshot.
  • path (String) — Snapshot file path on the disk.
  • disk_name (String) — Name of the disk that stores the snapshot.
  • size_bytes (UInt64) — Size of the snapshot file on disk.
  • last_modified_at (DateTime) — Last modification time of the snapshot file.
  • is_received (Bool) — true if the snapshot is currently being received from the leader. For such rows size_bytes and last_modified_at reflect the partial file as written so far and may underreport.
  • exists_on_disk (Bool) — Whether the snapshot file is currently present on disk. Always true for finalized snapshots (is_received = false) unless the file was removed out of band or is corrupted; may be false for snapshots being received (is_received = true) before any bytes are written.

Example:

SELECT * FROM system.keeper_snapshots ORDER BY last_log_index;
┌─last_log_index─┬─path──────────────────────────┬─disk_name─┬─size_bytes─┬────last_modified_at─┬─is_received─┬─exists_on_disk─┐
│           1000 │ snapshot_1000.bin.zstd        │ default   │      32468 │ 2026-05-22 14:00:00 │ false       │ true           │
│           2000 │ snapshot_2000.bin.zstd        │ default   │      48217 │ 2026-05-22 14:15:00 │ false       │ true           │
└────────────────┴───────────────────────────────┴───────────┴────────────┴─────────────────────┴─────────────┴────────────────┘
Navigation