elasticsearch/docs/reference/modules/cluster.asciidoc

[[modules-cluster]]
=== Cluster-level shard allocation and routing settings

include::{es-ref-dir}/modules/shard-allocation-desc.asciidoc[]

One of the main roles of the master is to decide which shards to allocate to
which nodes, and when to move shards between nodes in order to rebalance the
cluster.

There are a number of settings available to control the shard allocation process:

*   <<cluster-shard-allocation-settings>> control allocation and
    rebalancing operations.

*   <<disk-based-shard-allocation>> explains how Elasticsearch takes available
    disk space into account, and the related settings.

*   <<shard-allocation-awareness>> and <<forced-awareness>> control how shards
    can be distributed across different racks or availability zones.

*   <<cluster-shard-allocation-filtering>> allows certain nodes or groups of
    nodes excluded from allocation so that they can be decommissioned.

Besides these, there are a few other <<misc-cluster-settings,miscellaneous cluster-level settings>>.

include::cluster/shards_allocation.asciidoc[]

include::cluster/disk_allocator.asciidoc[]

[[shard-allocation-awareness-settings]]
==== Shard allocation awareness settings

You can use <<custom-node-attributes,custom node attributes>> as _awareness attributes_ to enable {es}
to take your physical hardware configuration into account when allocating shards.
If {es} knows which nodes are on the same physical server, in the same rack, or
in the same zone, it can distribute the primary shard and its replica shards to
minimize the risk of losing all shard copies in the event of a failure. <<shard-allocation-awareness,Learn more about shard allocation awareness>>.

`cluster.routing.allocation.awareness.attributes`::
      (<<dynamic-cluster-setting,Dynamic>>)
      The node attributes that {es} should use as awareness attributes. For example, if you have a `rack_id` attribute that specifies the rack in which each node resides, you can set this setting to `rack_id` to ensure that primary and replica shards are not allocated on the same rack. You can specify multiple attributes as a comma-separated list.

`cluster.routing.allocation.awareness.force.*`:: 
        (<<dynamic-cluster-setting,Dynamic>>)
        The shard allocation awareness values that must exist for shards to be reallocated in case of location failure. Learn more about <<forced-awareness,forced awareness>>.
      

include::cluster/allocation_filtering.asciidoc[]

include::cluster/misc.asciidoc[]