The `OPTIMIZE` query is supported for [MergeTree](../../engines/table-engines/mergetree-family/mergetree.md) family, the [MaterializedView](../../engines/table-engines/special/materializedview.md) and the [Buffer](../../engines/table-engines/special/buffer.md) engines. Other table engines aren’t supported.
When `OPTIMIZE` is used with the [ReplicatedMergeTree](../../engines/table-engines/mergetree-family/replication.md) family of table engines, ClickHouse creates a task for merging and waits for execution on all replicas (if the [alter_sync](../../operations/settings/settings.md#alter-sync) setting is set to `2`) or on current replica (if the [alter_sync](../../operations/settings/settings.md#alter-sync) setting is set to `1`).
- If `OPTIMIZE` does not perform a merge for any reason, it does not notify the client. To enable notifications, use the [optimize_throw_if_noop](../../operations/settings/settings.md#setting-optimize_throw_if_noop) setting.
- If you specify a `PARTITION`, only the specified partition is optimized. [How to set partition expression](alter/partition.md#how-to-set-partition-expression).
- If you specify `FINAL`, optimization is performed even when all the data is already in one part. You can control this behaviour with [optimize_skip_merged_partitions](../../operations/settings/settings.md#optimize-skip-merged-partitions). Also, the merge is forced even if concurrent merges are performed.
- If you specify `DEDUPLICATE`, then completely identical rows (unless by-clause is specified) will be deduplicated (all columns are compared), it makes sense only for the MergeTree engine.
You can specify how long (in seconds) to wait for inactive replicas to execute `OPTIMIZE` queries by the [replication_wait_for_inactive_replica_timeout](../../operations/settings/settings.md#replication-wait-for-inactive-replica-timeout) setting.
If the `alter_sync` is set to `2` and some replicas are not active for more than the time, specified by the `replication_wait_for_inactive_replica_timeout` setting, then an exception `UNFINISHED` is thrown.
If you want to perform deduplication on custom set of columns rather than on all, you can specify list of columns explicitly or use any combination of [`*`](../../sql-reference/statements/select/index.md#asterisk), [`COLUMNS`](../../sql-reference/statements/select/index.md#columns-expression) or [`EXCEPT`](../../sql-reference/statements/select/index.md#except-modifier) expressions. The explicitly written or implicitly expanded list of columns must include all columns specified in row ordering expression (both primary and sorting keys) and partitioning expression (partitioning key).
Notice that `*` behaves just like in `SELECT`: [MATERIALIZED](../../sql-reference/statements/create/table.md#materialized) and [ALIAS](../../sql-reference/statements/create/table.md#alias) columns are not used for expansion.
Also, it is an error to specify empty list of columns, or write an expression that results in an empty list of columns, or deduplicate by an `ALIAS` column.
When columns for deduplication are not specified, all of them are taken into account. Row is removed only if all values in all columns are equal to corresponding values in previous row:
When columns are specified implicitly, the table is deduplicated by all columns that are not `ALIAS` or `MATERIALIZED`. Considering the table above, these are `primary_key`, `secondary_key`, `value`, and `partition_key` columns:
Deduplicate by all columns that are not `ALIAS` or `MATERIALIZED` and explicitly not `value`: `primary_key`, `secondary_key`, and `partition_key` columns.