Elasticsearch — index/create

When to Use

Controlled schema definition: Define field types explicitly — keyword for exact-match filtering, text with a custom analyzer for full-text search, date for range queries — before any documents are indexed. Dynamic mapping can silently infer wrong types.
Custom analyzers: Configure language-specific analyzers, synonym filters, or n-gram tokenizers in the Settings block so search behaves exactly as intended from the first document.
Environment provisioning: A workflow provisions a new tenant or environment — create the required indices as part of setup before any data flows in.
Time-series index rotation: A scheduled workflow creates next month's log or metrics index (e.g. app-logs-2025-09) in advance, so the first write lands on a properly-configured index rather than triggering dynamic creation.

Dynamic mapping fallback: If you index a document into a non-existent index using document/create, Elasticsearch creates the index automatically with dynamic mapping. Use index/create first whenever field type control matters — particularly for fields used in aggregations, sorting, or exact-match queries, which require keyword mapping rather than text.

Configuration

Connection

Field	Required	Description
`Host`	Required	Elasticsearch cluster URL, e.g. `https://localhost:9200`.
`Username`	Optional	Elasticsearch username (required if security is enabled).
`Password`	Optional	Password for the specified username.
`IgnoreSSL`	Optional	Default `false`. Skip TLS validation. Development use only.

Operation

Field	Required	Description
`IndexName`	Required	Name for the new index. Must be lowercase with no spaces. Supports BizFirst expressions for dynamic names, e.g. `app-logs-{{ $now \| date('YYYY-MM') }}`.
`Mappings`	Optional	Elasticsearch mappings object defining field types and configurations. Entered as JSON in the code editor. If omitted, Elasticsearch uses dynamic mapping for all fields.
`Settings`	Optional	Elasticsearch index settings object — shard count, replica count, custom analyzers, refresh interval, etc. Entered as JSON. If omitted, Elasticsearch cluster defaults apply.

Validation Errors

Error Code	Cause
`VAL_MISSING_INDEX_NAME`	`IndexName` is empty.

Output

Success Port

Field	Type	Description
`status`	string	`"success"`
`indexName`	string	The name of the newly created index.
`created`	boolean	`true` when the index was created successfully.

Error Port

Fires if the index already exists (400 resource_already_exists_exception), authentication fails, or a settings/mappings validation error occurs. The error output contains errorCode and message fields.

Index already exists: If the index already exists, the operation fails and routes to the error port. Use index/get before this node to check existence, or handle the error port to treat "already exists" as acceptable in idempotent provisioning workflows.

Examples

Product Catalogue Index with Explicit Mappings

{
  "resource": "index",
  "operation": "create",
  "IndexName": "products",
  "Mappings": {
    "properties": {
      "title":       { "type": "text", "analyzer": "english" },
      "sku":         { "type": "keyword" },
      "category":    { "type": "keyword" },
      "brand":       { "type": "keyword" },
      "price":       { "type": "float" },
      "in_stock":    { "type": "boolean" },
      "indexed_at":  { "type": "date" },
      "tags":        { "type": "keyword" }
    }
  },
  "Settings": {
    "number_of_shards": 2,
    "number_of_replicas": 1
  }
}

sku, category, and brand use keyword so they support exact-match filters and aggregations. title uses the English analyzer for stemmed full-text search.

Time-Series Log Index — Dynamic Name

{
  "resource": "index",
  "operation": "create",
  "IndexName": "app-logs-{{ $now | date('YYYY-MM') }}",
  "Mappings": {
    "properties": {
      "timestamp":  { "type": "date" },
      "level":      { "type": "keyword" },
      "service":    { "type": "keyword" },
      "message":    { "type": "text" },
      "traceId":    { "type": "keyword" }
    }
  },
  "Settings": {
    "number_of_shards": 1,
    "number_of_replicas": 0,
    "refresh_interval": "5s"
  }
}

Run this via a scheduled workflow at the start of each month. The refresh_interval of 5s reduces indexing overhead for high-volume log ingestion.

Index with Custom Analyzer

{
  "resource": "index",
  "operation": "create",
  "IndexName": "articles",
  "Settings": {
    "analysis": {
      "analyzer": {
        "autocomplete": {
          "tokenizer": "autocomplete_tokenizer",
          "filter": ["lowercase"]
        }
      },
      "tokenizer": {
        "autocomplete_tokenizer": {
          "type": "edge_ngram",
          "min_gram": 2,
          "max_gram": 20,
          "token_chars": ["letter", "digit"]
        }
      }
    }
  },
  "Mappings": {
    "properties": {
      "title": {
        "type": "text",
        "analyzer": "autocomplete",
        "search_analyzer": "standard"
      }
    }
  }
}

Configures an edge n-gram tokenizer for autocomplete search on the title field. The search_analyzer uses standard so query terms are not n-grammed on the query side.

index/create FormID 20300

When to Use

Configuration

Connection

Operation

Validation Errors

Output

Success Port

Error Port

Examples

Product Catalogue Index with Explicit Mappings

Time-Series Log Index — Dynamic Name

Index with Custom Analyzer