> ## Documentation Index
> Fetch the complete documentation index at: https://private-7c7dfe99-fix-nav-issues.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

> System table which exists only when this node runs an in-process ClickHouse Keeper. Contains one row per Raft cluster member as seen by this Keeper.

# system.keeper_cluster

<Info>
  **Querying in ClickHouse Cloud**

  The data in this system table is held locally on each node in ClickHouse Cloud. Obtaining a complete view of all data, therefore, requires the `clusterAllReplicas` function. See [here](/reference/system-tables/overview#system-tables-in-clickhouse-cloud) for further details.
</Info>

<h2 id="description">
  Description
</h2>

This table does not exist if this node is not configured to run an in-process ClickHouse Keeper. It contains one row per Raft cluster member, fusing static cluster topology (from the Raft configuration) with the local node's own log position.

Every node fills exactly one `last_log_index` value — the row matching its own `server_id`. Peer log positions are not surfaced here because they are tracked only on the leader and that view is not symmetric across the cluster.

<h2 id="columns">
  Columns
</h2>

* `server_id` ([Int32](/reference/data-types/int-uint)) — Raft server id of this cluster member.
* `host` ([String](/reference/data-types/string)) — Host parsed from the endpoint (prefix before the last colon).
* `endpoint` ([String](/reference/data-types/string)) — Raw Raft endpoint as configured (`host:port`).
* `is_observer` ([Bool](/reference/data-types/boolean)) — `true` if this member is a non-voting observer, configured via `<can_become_leader>false</can_become_leader>`.
* `priority` ([Int32](/reference/data-types/int-uint)) — Raft priority of this member; higher values are preferred during leader election.
* `is_leader` ([Bool](/reference/data-types/boolean)) — `true` if this member is the current Raft leader.
* `is_self` ([Bool](/reference/data-types/boolean)) — `true` if this row describes the local Keeper node.
* `last_log_index` ([Nullable(UInt64)](/reference/data-types/int-uint)) — Highest log index in the local Raft log store on this node. Populated only on the row matching the current node (`is_self = true`); `NULL` for other rows.

Example:

```sql theme={null}
SELECT * FROM system.keeper_cluster ORDER BY server_id;
```

```text theme={null}
┌─server_id─┬─host──┬─endpoint───┬─is_observer─┬─priority─┬─is_leader─┬─is_self─┬─last_log_index─┐
│         1 │ node1 │ node1:9234 │ false       │        3 │ true      │ true    │             42 │
│         2 │ node2 │ node2:9234 │ false       │        2 │ false     │ false   │           ᴺᵁᴸᴸ │
│         3 │ node3 │ node3:9234 │ true        │        1 │ false     │ false   │           ᴺᵁᴸᴸ │
└───────────┴───────┴────────────┴─────────────┴──────────┴───────────┴─────────┴────────────────┘
```