Shards Allocation

Shards allocation is the process of allocating shards to nodes. This can happen during initial recovery, replica allocation, rebalancing, or handling nodes being added or removed.

Shard Allocation Awareness

Cluster allocation awareness allows to configure shard and replicas allocation across generic attributes associated the nodes. Lets explain it through an example:

Assume we have several racks. When we start a node, we can configure an attribute called rack_id (any attribute name works), for example, here is a sample config:

node.rack_id: rack_one

The above sets an attribute called rack_id for the relevant node with a value of rack_one. Now, we need to configure the rack_id attribute as one of the awareness allocation attributes (set it on all (master eligible) nodes config):

cluster.routing.allocation.awareness.attributes: rack_id

The above will mean that the rack_id attribute will be used to do awareness based allocation of shard and its replicas. For example, lets say we start 2 nodes with node.rack_id set to rack_one, and deploy a single index with 5 shards and 1 replica. The index will be fully deployed on the current nodes (5 shards and 1 replica each, total of 10 shards).

Now, if we start two more nodes, with node.rack_id set to rack_two, shards will relocate to even the number of shards across the nodes, but, a shard and its replica will not be allocated in the same rack_id value.

The awareness attributes can hold several values, for example:

cluster.routing.allocation.awareness.attributes: rack_id,zone

NOTE: When using awareness attributes, shards will not be allocated to nodes that don’t have values set for those attributes.

Forced Awareness

Sometimes, we know in advance the number of values an awareness attribute can have, and more over, we would like never to have more replicas then needed allocated on a specific group of nodes with the same awareness attribute value. For that, we can force awareness on specific attributes.

For example, lets say we have an awareness attribute called zone, and we know we are going to have two zones, zone1 and zone2. Here is how we can force awareness one a node:

cluster.routing.allocation.awareness.force.zone.values: zone1,zone2
cluster.routing.allocation.awareness.attributes: zone

Now, lets say we start 2 nodes with node.zone set to zone1 and create an index with 5 shards and 1 replica. The index will be created, but only 5 shards will be allocated (with no replicas). Only when we start more shards with node.zone set to zone2 will the replicas be allocated.

Automatic Preference When Searching / GETing

When executing a search, or doing a get, the node receiving the request will prefer to execute the request on shards that exists on nodes that have the same attribute values as the executing node.

Realtime Settings Update

The settings can be updated using the cluster update settings API on a live cluster.

Shard Allocation Filtering

Allow to control allocation if indices on nodes based on include/exclude filters. The filters can be set both on the index level and on the cluster level. Lets start with an example of setting it on the cluster level:

Lets say we have 4 nodes, each has specific attribute called tag associated with it (the name of the attribute can be any name). Each node has a specific value associated with tag. Node 1 has a setting node.tag: value1, Node 2 a setting of node.tag: value2, and so on.

We can create an index that will only deploy on nodes that have tag set to value1 and value2 by setting index.routing.allocation.include.tag to value1,value2. For example:

curl -XPUT localhost:9200/test -d '{
    "index.routing.allocation.include.tag" : "value1,value2"
}'

On the other hand, we can create an index that will be deployed on all nodes except for nodes with a tag of value value3 by setting index.routing.allocation.exclude.tag to value3. For example:

curl -XPUT localhost:9200/test -d '{
    "index.routing.allocation.exclude.tag" : "value3"
}'

The include and exclude values can have generic simple matching wildcards, for example, value1*. A special attribute name called _ip can be used to match on node ip values.

Obviously a node can have several attributes associated with it, and both the attribute name and value are controlled in the setting. For example, here is a sample of several node configurations:

node.group1: group1_value1
node.group2: group2_value4

In the same manner, include and exclude can work against several attributes, for example:

curl -XPUT localhost:9200/test -d '{
    "index.routing.allocation.include.group1" : "xxx"
    "index.routing.allocation.include.group2" : "yyy",
    "index.routing.allocation.exclude.group3" : "zzz",
}'

The provided settings can also be updated in real time using the update settings API, allowing to “move” indices (shards) around in realtime.

Cluster wide filtering can also be defined, and be updated in real time using the cluster update settings API. This setting can come in handy for things like decommissioning nodes (even if the replica count is set to 0). Here is a sample of how to decommission a node based on _ip address:

curl -XPUT localhost:9200/_cluster/_settings -d '{
    "transient" : {
        "cluster.routing.allocation.exclude._ip" : "10.0.0.1"
    }
}'