elasticsearch/docs/reference/modules/gateway.asciidoc

[[modules-gateway]]
=== Local gateway settings

[[dangling-indices]]
The local gateway stores the cluster state and shard data across full
cluster restarts.

The following _static_ settings, which must be set on every <<master-node-role,master-eligible node>>,
control how long a freshly elected master should wait before it tries to
recover the <<cluster-state,cluster state>> and the cluster's data.

NOTE: These settings only take effect during a <<restart-cluster-full,full cluster restart>>.

`gateway.expected_data_nodes`::
(<<static-cluster-setting,Static>>)
Number of data nodes expected in the cluster.
Recovery of local shards begins when the expected number of
data nodes join the cluster. Defaults to `0`.

`gateway.recover_after_time`::
(<<static-cluster-setting,Static>>)
If the expected number of nodes is not achieved, the recovery process waits
for the configured amount of time before trying to recover.
Defaults to `5m`.
+
Once the `recover_after_time` duration has timed out, recovery will start
as long as the following condition is met:

`gateway.recover_after_data_nodes`::
(<<static-cluster-setting,Static>>)
Recover as long as this many data nodes have joined the cluster.

These settings can be configured in `elasticsearch.yml` as follows:

[source,yaml]
--------------------------------------------------
gateway.expected_data_nodes: 3
gateway.recover_after_time: 600s
gateway.recover_after_data_nodes: 3
--------------------------------------------------