Fielddata in Elasticsearch

What it is:

In Elasticsearch the term Fielddata is relevant when sorting and doing aggregations (similar to SQL GROUP BY COUNT and AVERAGE functions) on text fields.  

For performance reasons, there are some rules as to the kinds of fields that can be aggregated. You can group by any numeric field but for text fields, which have to be of keyword type or have fielddata=true since they don’t support doc_values (Doc values are the on-disk inverted index data structure, built at document indexing time, which makes aggregations possible). 

Fielddata is an in-memory data structure used by text fields for the same purpose. Since it uses a lot of heap size it is disabled by default.

Examples:

The following PUT mapping API call will enable Fielddata on my_field text field.

PUT my_index/_mapping{"properties":{"my_field":{"type":"text","fielddata":true}}}

Notes:

• As field-data is disabled by default on text fields, in case of an attempt  to aggregate on a text field with field-data disabled, you would get the following error message:
  “Fielddata is disabled on text fields by default.  Set `fielddata=true` on [`your_field_name`] in order to load field data in memory by uninverting the inverted index. Note that this can however, use “significant memory.” – if this happens you can either enable the field-data on that text field, or choose another way to query the data (again, because field-data consumes a lot of memory and is not recommended).

Index in Elasticsearch

What it is:

In Elasticsearch, an index (plural: indices) can be thought of as a table inside a database. An index contains a schema and can have one or more shards and replicas. An Elasticsearch index is divided into shards and each shard is an instance of a Lucene index.

Indices are used to store the documents in dedicated data structures corresponding to the data type of fields. For example, text fields are stored inside an inverted index whereas numeric and geo fields are stored inside BKD trees.

Examples:

Create index:

The following example is based on Elasticsearch version 5.x onwards. An index with two shards, each having one replica will be created with the name test_index1

PUT /test_index1?pretty
{
    "settings" : {
        "number_of_shards" : 2,
        "number_of_replicas" : 1
    },
    "mappings" : {
        "properties" : {
            "tags" : { "type" : "keyword" },
            "updated_at" : { "type" : "date" }
        }
    }
}

List indices:

All the index names and their basic information can be retrieved using the following command:

GET _cat/indices?v

Index a document:

Let’s add a document in the index with below command:

PUT test_index1/_doc/1
{
  "tags": [
    "opster",
    "elasticsearch"
  ],
  "date": "01-01-2020"
}

Query an index:

GET test_index1/_search
{
  "query": {
    "match_all": {}
  }
}

Query multiple indices:

It is possible to search multiple indices with a single request. If it is a raw HTTP request, index names should be sent in comma-separated format, as shown in the example below, and in the case of a query via a programming language client such as python or Java, index names are to be sent in a list format.

GET test_index1,test_index2/_search

Delete indices:

DELETE test_index1

Common problems:

DELETE /*

action.destructive_requires_name: true

Memory in Elasticsearch

What is it:

Memory is one of the most critical resources to monitor in Elasticsearch. Elasticsearch runs on JVM and uses heap memory areas for query cache, request cache, accessing lucene segments and storing fielddata for aggregations and sorting.

Common problems and important points:

• The most common error that arises in Elasticsearch is OutOfMemory error. This error comes when the node is not able to cope up with the required heap size space. To avoid this, you need to closely monitor the heap utilization and garbage collector performance.
• As per the most up-to-date best practices you should not allocate more than 50 percent of total RAM to JVM heap size. Starting from Elasticsearch version 5.x onward this can be set using -Xms and -Xmx parameters inside jvm.options configuration file. The defaults are set to 1 GB for both minimum and maximum heap size.
• The heap size should not set more than 31 GB in any case to avoid the poor garbage collection.