{
"schema_version": 2,
"id": "operate/rs/references/rest-api/objects/cluster_settings",
"title": "Cluster settings object",
"url": "https://redis.io/docs/latest/operate/rs/references/rest-api/objects/cluster_settings/",
"summary": "An object for cluster resource management settings",
"tags": [
"docs",
"operate",
"rs"
],
"last_updated": "2026-02-19T10:49:21-06:00",
"page_type": "content",
"content_hash": "ecde749c771887b4f488d1fefadf5bd8154df9c2264bb5ff2a017d22601747b8",
"sections": [
{
"id": "content",
"title": "Content",
"role": "content",
"text": "Cluster resources management policy\n\n| Name | Type/Value | Description |\n|------|------------|-------------|\n| acl_pubsub_default | `resetchannels`
`allchannels` | Default pub/sub ACL rule for all databases in the cluster:
•`resetchannels` blocks access to all channels (restrictive)
•`allchannels` allows access to all channels (permissive) |\n| auto_recovery | boolean (default: false) | Defines whether to use automatic recovery after shard failure |\n| automatic_node_offload | boolean (default: true) | Defines whether the cluster will automatically migrate shards from a node, in case the node is overbooked |\n| bigstore_migrate_node_threshold | integer | Minimum free memory (excluding reserved memory) allowed on a node before automatic migration of shards from it to free more memory |\n| bigstore_migrate_node_threshold_p | integer | Minimum free memory (excluding reserved memory) allowed on a node before automatic migration of shards from it to free more memory |\n| bigstore_provision_node_threshold | integer | Minimum free memory (excluding reserved memory) allowed on a node before new shards can no longer be added to it |\n| bigstore_provision_node_threshold_p | integer | Minimum free memory (excluding reserved memory) allowed on a node before new shards can no longer be added to it |\n| data_internode_encryption | boolean | Enable/deactivate encryption of the data plane internode communication |\n| db_conns_auditing | boolean | [Audit connections](https://redis.io/docs/latest/operate/rs/security/audit-events) for new databases by default if set to true. |\n| default_bigstore_version | **`1`**
`2` | Determines the default value of `bigstore_version` for new databases with Redis version 7.4:
• `1` for Auto Tiering (Redis on Flash version 1). Default version.
• `2` for Redis Flex (Redis on Flash version 2) on databases that support it. Can only be used with the `speedb` driver. |\n| default_concurrent_restore_actions | integer | Default number of restore actions allowed at the same time. Set to 0 to allow any number of simultaneous restore actions. |\n| default_fork_evict_ram | boolean | If true, the bdbs should evict data from RAM to ensure successful replication or persistence |\n| default_non_sharded_proxy_policy | `single`
`all-master-shards`
`all-nodes` | Default proxy_policy for newly created non-sharded databases' endpoints |\n| default_oss_sharding | boolean (default: false) | Default hashing policy to use for new databases. This field is for future use only and should not be changed. |\n| default_oss_cluster | boolean (default: false) | Default `oss_cluster` setting for new databases. Enables the OSS Cluster API if `true`, or turns it off if `false`. |\n| default_provisioned_redis_version | string | Default Redis version |\n| default_recovery_wait_time | integer (default: -1) | The default time for new databases to wait for the persistence file to be available during automatic recovery. -1 means wait forever. |\n| default_sharded_proxy_policy | `single`
`all-master-shards`
`all-nodes` | Default proxy_policy for newly created sharded databases' endpoints |\n| default_shards_placement | `dense`
`sparse` | Default shards_placement for a newly created databases |\n| default_tracking_table_max_keys_policy | integer (default: 1000000) | Defines the default value of the client-side caching invalidation table size for new databases. 0 makes the cache unlimited. |\n| diskless_repl | `yes`
`no` (default: yes) | If `yes`, enable default Redis diskless replication mechanism (deprecated; use `repl_diskless` instead) |\n| endpoint_rebind_propagation_grace_time | integer | Time to wait between the addition and removal of a proxy |\n| evict_node_use_free_memory | boolean | When evicting a node, use the free memory instead of the provisional memory to check if the shards from the old node fit on the new one |\n| expose_hostnames_for_all_suffixes | boolean (default: false) | If true, enables exposing hostnames for non-default DNS suffixes |\n| failure_detection_sensitivity | `high`
`low` | Predefined thresholds and timeouts for failure detection (previously known as `watchdog_profile`)
• `high` (previously `local-network`) – high failure detection sensitivity, lower thresholds, faster failure detection and failover
• `low` (previously `cloud`) – low failure detection sensitivity, higher tolerance for latency variance (also called network jitter) |\n| filter_node_wd_health_metrics | boolean (default: true) | When enabled, `node_wd` health metrics (`shard_ping_failures`, `endpoint_ping_failures`, `dmc_ping_failures` and their duration counterparts) are not emitted to `statsd` |\n| hide_user_data_from_log | boolean (default: false) | Set to `true` to enable the `hide-user-data-from-log` Redis configuration setting, which avoids logging user data |\n| login_lockout_counter_reset_after | integer | Number of seconds that must elapse between failed sign in attempts before the lockout counter is reset to 0. |\n| login_lockout_duration | integer | Duration (in secs) of account lockout. If set to 0, the account lockout will persist until released by an admin. |\n| login_lockout_threshold | integer | Number of failed sign in attempts allowed before locking a user account |\n| master_healthcheck_api_auth | boolean (default: true) | Defines if authentication is required by the local `master_healthcheck` API |\n| max_redis_forks | integer (default: 0) | Maximum number of background processes forked from shards that can exist on the node at any given time. 0 means unlimited. |\n| max_saved_events_per_type | integer | Maximum saved events per event type |\n| max_simultaneous_backups | integer (default: 4) | Maximum number of backup processes allowed at the same time |\n| metrics_exporter_expose_bdb_name | boolean (default: false) | If true, adds a label with the database name to relevant metrics |\n| parallel_shards_upgrade | integer | Maximum number of shards to upgrade in parallel |\n| max_slave_full_syncs | integer (default: 0) | Maximum number of simultaneous replica full syncs that can run at any given time. 0 means unlimited. |\n| oss_cluster_api_preferred_endpoint_type | `ip`
`hostname` (default: ip) | Determines the default endpoint type in the OSS Cluster API for new endpoints |\n| oss_cluster_api_preferred_ip_type | `internal`
`external` (default: internal) | Determines the default IP type in the OSS Cluster API for new endpoints |\n| persistence_cleanup_grace_time | integer | Time in seconds before an unmodified file is considered to be stale and to be removed by persistence cleanup |\n| persistence_cleanup_scan_interval | string | [CRON expression](https://en.wikipedia.org/wiki/Cron#CRON_expression) that defines the Redis cleanup schedule |\n| persistent_node_removal | boolean | When removing a node, wait for persistence files to be created for all migrated shards |\n| rack_aware | boolean | Cluster operates in a rack-aware mode |\n| redis_migrate_node_threshold | integer | Minimum free memory (excluding reserved memory) allowed on a node before automatic migration of shards from it to free more memory |\n| redis_migrate_node_threshold_p | integer | Minimum free memory (excluding reserved memory) allowed on a node before automatic migration of shards from it to free more memory |\n| redis_provision_node_threshold | integer | Minimum free memory (excluding reserved memory) allowed on a node before new shards can no longer be added to it |\n| redis_provision_node_threshold_p | integer | Minimum free memory (excluding reserved memory) allowed on a node before new shards can no longer be added to it |\n| redis_upgrade_policy | **`major`**
`latest` | Create/upgrade Redis Software on databases in the cluster by compatibility with major versions or latest versions of Redis Open Source |\n| repl_diskless | boolean (default: true) | If true, enables the default Redis diskless replication mechanism |\n| resp3_default | boolean (default: true) | Determines the default value of the `resp3` option upon upgrading a database to version 7.2 |\n| shards_overbooking | boolean | If true, all databases' memory_size is ignored during shards placement |\n| show_internals | boolean | Show internal databases (and their shards and endpoints) REST APIs |\n| show_metrics_during_state_machine | boolean | Show metrics during state machine operations |\n| slave_ha | boolean | Enable the replica high-availability mechanism. Deprecated as of Redis Software v7.2.4. |\n| slave_ha_bdb_cooldown_period | integer | Time in seconds between runs of the replica high-availability mechanism on different nodes on the same database |\n| slave_ha_cooldown_period | integer | Time in seconds between runs of the replica high-availability mechanism on different nodes on the same database |\n| slave_ha_grace_period | integer | Time in seconds between a node failure and when the replica high-availability mechanism starts relocating shards |\n| slow_log_max_len | integer (default: 1024) | Set max slow log entries in debug info |\n| use_librdb | boolean | If true, new databases use the new RDB parser instead of the old one |\n| witness_disk_update_frequency_divisor | integer (default: 3) | Redis Software updates the witness disk every `node_max_update_time / witness_disk_update_frequency_divisor` seconds |"
}
],
"examples": []
}