elastic-search.md

elastic-search

• aur/elasticsearch

• aur/kibana-bin

• installation

# first time
elasticsearch-keystore create
# the above command will create `/etc/elasticsearch/elasticsearch.keystore` file

# service
systemctl status elasticsearch
systemctl enable elasticsearch
systemctl start elasticsearch

# I don't know where to find the default password for `elastic` user
# just run below command to reset the password for `elastic` user
elasticsearch-reset-password -u elastic
# this will prompt for new password

• documentation: https://www.elastic.co/guide/en/elasticsearch/reference/current/docs.html

• shards are for indices

  • split api - increase shard count
  • shrink api - decrease shard count

• replication

  • while creating index we can choose replica count
  • a replica is a copy of a shard

basic

• view info
  • cluster: GET {{local}}/_cluster/health
  • node: GET {{local}}/_cat/nodes?v=true
  • index: GET {{local}}/_cat/indices?v=true
• get all documents

  GET {{local}}/students/_search
  Content-Type: application/json
  
  {
    "query": {
      "match_all": {}
    }
  }

create new index

PUT  {{local}}/students
Content-Type: application/json

{
	"settings": {
		"number_of_shards": "3",
		"number_of_replicas": "3"
	}
}

adding a document

POST {{local}}/students/_doc
Content-Type: application/json

{
	"fullname": "Foo Bar",
	"dob": "2000-01-01",
	"gender": "male",
	"address": "aaaaaaaa bbbbbbb",
	"contact": "297234923498"
}

updating single field (its more complext than what I thought)

POST {{local}}/students/_update/28
Content-Type: application/json

{
	"script": {
		"source": "ctx._source.address=(params.newAddress)",
		"lang": "painless",
		"params": {
			"newAddress": "ssssssssss sssssssssss"
		}
	}
}

update single field - simple way

POST {{local}}/students/_update/28
Content-Type: application/json

{
	"doc": {
		"address": "ssssssssss ttttttttttttttt"
	}
}

creating mapping

PUT /students
Content-Type: application/json

{
  "mapping": {
    "properties": {
      "fullname": { "type": "text" },
      "dob": { "type": "date", "format": "strict_date_optional_time||epoch_millis" },
      "gender": { "type": "text" },
      "address": { "type": "text" },
      "email": { "type": "keyword" }
    }
  }
}

get the mapping

# get mapping of an index
GET /students/_mapping

# get mapping of one field
GET /students/_mapping/field/address

updating existing mapping

adding new key to exiting mapping

PUT /students/_mapping
Content-Type: application/json

{
  "properties": {
    "class": { "type": "integer" },
    "report": {
      "type": "nested"
      "properties": {
        "class": { "type": "integer", "coerce": false },
        "percentage": { "type": "float" }
      }
    }
  }
}

## mapping options

- `"norms": false` - normaliztion, ranking; can be disabled if ranking not important for a field
- `"doc_values": false` - enabled by default, it should be enabled for aggregate operations. disabling increases indexing speed.
- `"coerce": false` - enable/disable automatic type coercion (enabled by default)
- `"index": false` - exclude a field from searching
- `"null_value": "NULL" - this is the default value should be used when document created with null value for this field. ES by default ignores null values & does not index it.