@@ -43,8 +43,8 @@ to be allocated to a node. This is in contrast to `include` which will
43
43
include a node if ANY rule matches.
44
44
45
45
The `include`, `exclude` and `require` values can have generic simple
46
- matching wildcards, for example, `value1*`. Additonally, special attribute
47
- names called `_ip`, `_name`, `_id` and `_host` can be used to match by node
46
+ matching wildcards, for example, `value1*`. Additonally, special attribute
47
+ names called `_ip`, `_name`, `_id` and `_host` can be used to match by node
48
48
ip address, name, id or host name, respectively.
49
49
50
50
Obviously a node can have several attributes associated with it, and
@@ -100,16 +100,18 @@ settings API.
100
100
[[disk]]
101
101
=== Disk-based Shard Allocation
102
102
103
+ coming[1.3.0] disk based shard allocation is enabled from version 1.3.0 onward
104
+
103
105
Elasticsearch can be configured to prevent shard
104
106
allocation on nodes depending on disk usage for the node. This
105
- functionality is disabled by default, and can be changed either in the
107
+ functionality is enabled by default, and can be changed either in the
106
108
configuration file, or dynamically using:
107
109
108
110
[source,js]
109
111
--------------------------------------------------
110
112
curl -XPUT localhost:9200/_cluster/settings -d '{
111
113
"transient" : {
112
- "cluster.routing.allocation.disk.threshold_enabled" : true
114
+ "cluster.routing.allocation.disk.threshold_enabled" : false
113
115
}
114
116
}'
115
117
--------------------------------------------------
@@ -118,15 +120,15 @@ Once enabled, Elasticsearch uses two watermarks to decide whether
118
120
shards should be allocated or can remain on the node.
119
121
120
122
`cluster.routing.allocation.disk.watermark.low` controls the low
121
- watermark for disk usage. It defaults to 70 %, meaning ES will not
122
- allocate new shards to nodes once they have more than 70 % disk
123
+ watermark for disk usage. It defaults to 85 %, meaning ES will not
124
+ allocate new shards to nodes once they have more than 85 % disk
123
125
used. It can also be set to an absolute byte value (like 500mb) to
124
126
prevent ES from allocating shards if less than the configured amount
125
127
of space is available.
126
128
127
129
`cluster.routing.allocation.disk.watermark.high` controls the high
128
- watermark. It defaults to 85 %, meaning ES will attempt to relocate
129
- shards to another node if the node disk usage rises above 85 %. It can
130
+ watermark. It defaults to 90 %, meaning ES will attempt to relocate
131
+ shards to another node if the node disk usage rises above 90 %. It can
130
132
also be set to an absolute byte value (similar to the low watermark)
131
133
to relocate shards once less than the configured amount of space is
132
134
available on the node.
0 commit comments