This post explains the installation steps of ElasticSearch(1.3.4) cluster on linux machine. For ElasticSeach, It is required to have java installed on machines. If java is not available on machines
then first install it.
After
installing the java on machines, follow the below steps to install the ElasticSearch.
1. Download ElasticSearch: Download the ElasticSearch bundle from the website and extract the bundle using the below commands.
wget https://download.elasticsearch.org/elasticsearch/elasticsearch/elasticsearch-1.3.4.zip
tar –xvf elasticsearch-1.3.4.zip
3. Configure ElasticSearch: To configure the ElasticSearch,
It is required to add the following properties in elasticsearch.yaml file, which is available inside the conf directory of extracted directory.
a. cluster.name: This is the name of elastic search cluster and it should be same across all nodes of the cluster. This property is responsible for joining all the nodes in cluster.
b. node.name: This property defines the name of node. It should be unique across all the nodes.
c. path.data: This property defines the path for the data which is going to stored in elastic search for the indexes.
d. action.auto_create_index: This property restricts the auto creation of indexes. It accepts the true/false value.
e. index.number_of_shards: It is used to set the number of shards for the index.
f. index.number_of_replicas: It is used to set the number of replication for the index.
g. bootstrap.mlockall: It Locks the memory for better performance of ElasticSearch.
a. cluster.name: This is the name of elastic search cluster and it should be same across all nodes of the cluster. This property is responsible for joining all the nodes in cluster.
b. node.name: This property defines the name of node. It should be unique across all the nodes.
c. path.data: This property defines the path for the data which is going to stored in elastic search for the indexes.
d. action.auto_create_index: This property restricts the auto creation of indexes. It accepts the true/false value.
e. index.number_of_shards: It is used to set the number of shards for the index.
f. index.number_of_replicas: It is used to set the number of replication for the index.
g. bootstrap.mlockall: It Locks the memory for better performance of ElasticSearch.
4. Start ElasticSearch: To start the
elastic search run the below command from the bin directory of extracted
elastic search directory. A flag -d is used to run the elasticsearch in daemon
mode.
./elasticsearch -d
ElasticSearch Cluster Setup on two nodes:
To install the elastic search as a cluster, you have to perform the above 3 steps on the nodes where you want to install the elastic search cluster. To configure the ElasticSearch let’s take an example of 2 nodes for the elastic search cluster.
ElasticSearch Configuration on node1 in elasticsearch.yaml file.
cluster.name : elasticsearch_cluster
path.data : /home/$username/elasticsearch-1.3.4/data_dir
action.auto_create_index : false
index.number_of_shards : 2
index.number_of_replicas : 1
bootstrap.mlockall : false
ElasticSearch configuration on node2 in elasticsearch.yaml file.
node.name : node2
path.data : /home/$username/elasticsearch-1.3.4/data_dir
action.auto_create_index : false
index.number_of_shards : 2
index.number_of_replicas : 1
bootstrap.mlockall : false
Now run the elastic search on both the nodes using the command mentioned in 3rd step.
