elasticsearch - moving from multi servers to one server - elasticsearch

Question

I have a cluster of 5 servers for elasticsearch, all with the same version of elasticsearch.
I need to move all data from servers 2, 3, 4, 5 to server 1.
How can I do it?
How can I know which server has data at all?
After change of _cluster/settings with:
PUT _cluster/settings
{
"persistent" : {
"cluster.routing.allocation.require._host" : "server1"
}
}
I get for: curl -GET http://localhost:9200/_cat/allocation?v
the following:
shards disk.indices disk.used disk.avail disk.total disk.percent host ip node
6 54.5gb 170.1gb 1.9tb 2.1tb 7 *.*.*.* *.*.*.* node-5
6 50.4gb 167.4gb 1.9tb 2.1tb 7 *.*.*.* *.*.*.* node-3
6 22.6gb 139.8gb 2tb 2.1tb 6 *.*.*.* *.*.*.* node-2
6 49.8gb 166.6gb 1.9tb 2.1tb 7 *.*.*.* *.*.*.* node-4
6 54.8gb 172.1gb 1.9tb 2.1tb 7 *.*.*.* *.*.*.* node-1
and for: GET _cluster/settings?include_defaults
the following:
#! Deprecation: [node.max_local_storage_nodes] setting was deprecated in Elasticsearch and will be removed in a future release!
{
"persistent" : {
"cluster" : {
"routing" : {
"allocation" : {
"require" : {
"_host" : "server1"
}
}
}
}
},
"transient" : { },
"defaults" : {
"cluster" : {
"max_voting_config_exclusions" : "10",
"auto_shrink_voting_configuration" : "true",
"election" : {
"duration" : "500ms",
"initial_timeout" : "100ms",
"max_timeout" : "10s",
"back_off_time" : "100ms",
"strategy" : "supports_voting_only"
},
"no_master_block" : "write",
"persistent_tasks" : {
"allocation" : {
"enable" : "all",
"recheck_interval" : "30s"
}
},
"blocks" : {
"read_only_allow_delete" : "false",
"read_only" : "false"
},
"remote" : {
"node" : {
"attr" : ""
},
"initial_connect_timeout" : "30s",
"connect" : "true",
"connections_per_cluster" : "3"
},
"follower_lag" : {
"timeout" : "90000ms"
},
"routing" : {
"use_adaptive_replica_selection" : "true",
"rebalance" : {
"enable" : "all"
},
"allocation" : {
"node_concurrent_incoming_recoveries" : "2",
"node_initial_primaries_recoveries" : "4",
"same_shard" : {
"host" : "false"
},
"total_shards_per_node" : "-1",
"shard_state" : {
"reroute" : {
"priority" : "NORMAL"
}
},
"type" : "balanced",
"disk" : {
"threshold_enabled" : "true",
"watermark" : {
"low" : "85%",
"flood_stage" : "95%",
"high" : "90%"
},
"include_relocations" : "true",
"reroute_interval" : "60s"
},
"awareness" : {
"attributes" : [ ]
},
"balance" : {
"index" : "0.55",
"threshold" : "1.0",
"shard" : "0.45"
},
"enable" : "all",
"node_concurrent_outgoing_recoveries" : "2",
"allow_rebalance" : "indices_all_active",
"cluster_concurrent_rebalance" : "2",
"node_concurrent_recoveries" : "2"
}
},
...
"nodes" : {
"reconnect_interval" : "10s"
},
"service" : {
"slow_master_task_logging_threshold" : "10s",
"slow_task_logging_threshold" : "30s"
},
...
"name" : "cluster01",
...
"max_shards_per_node" : "1000",
"initial_master_nodes" : [ ],
"info" : {
"update" : {
"interval" : "30s",
"timeout" : "15s"
}
}
},
...

Answer 1

You can use shard allocation filtering to move all your data to server 1.
Simply run this:
PUT _cluster/settings
{
"persistent" : {
"cluster.routing.allocation.require._name" : "node-1",
"cluster.routing.allocation.exclude._name" : "node-2,node-3,node-4,node-5"
}
}
Instead of _name you can also use _ip or _host depending on what is more practical for you.
After running this command, all primary shards will migrate to server1 (the replicas will be unassigned). You just need to make sure that server1 has enough storage space to store all the primary shards.
If you want to get rid of the unassigned replicas (and get back to green state), simply run this:
PUT _all/_settings
{
"index" : {
"number_of_replicas" : 0
}
}