Mapping & Index Management

Inspect the search index behind a Dataset, and administratively remove orphaned indices.

GET /api/datasets/{id}/mapping

Get the index mapping for the Dataset.

Authentication: Required

Path Parameters:

Parameter	Type	Description
id	string	Dataset UUID or ownerId:slug natural id

Get the index mappingGET /api/datasets/admin:inventory-levels/mapping

The raw Elasticsearch mapping, keyed by index name. Answers {} when the Dataset has never been indexed.

Response · 200

{
  "_links": {
    "self": {
      "href": "https://informer.example.com/api/datasets/admin%3Ainventory-levels/mapping"
    }
  },
  "_meta": {
    "datasetId": "00000000-0000-4000-8000-000000000001",
    "systemId": "00000000-0000-4000-8000-000000000002"
  },
  "dynamic_templates": [
    {
      "string_as_keyword": {
        "match_mapping_type": "string",
        "mapping": {
          "fields": {
            "text": {
              "type": "text"
            }
          },
          "ignore_above": 256,
          "type": "keyword"
        }
      }
    },
    {
      "double_to_float": {
        "match_mapping_type": "double",
        "mapping": {
          "type": "float"
        }
      }
    },
    {
      "long_to_float": {
        "match_mapping_type": "long",
        "mapping": {
          "type": "float"
        }
      }
    }
  ],
  "properties": {
    "onHand": {
      "type": "float"
    },
    "product": {
      "type": "keyword",
      "ignore_above": 256,
      "fields": {
        "text": {
          "type": "text"
        }
      }
    },
    "reorderAt": {
      "type": "float"
    },
    "sku": {
      "type": "keyword",
      "ignore_above": 256,
      "fields": {
        "text": {
          "type": "text"
        }
      }
    }
  }
}

Captured from the API examples test suite; ids and timestamps are normalized.

Key Parts:

Section	Description
_meta	Informer's ownership stamp on the index (datasetId, systemId)
dynamic_templates	Informer's default type coercions (strings index as keyword + .text sub-field; numerics as float)
properties	Per-field index types

Answers {} when the Dataset has never been indexed.

Use Case:

Debugging search behavior: the mapping is the ground truth for how each field is indexed, sorted, and searched.

---

GET /api/datasets/{id}/index/status

Get the index's shard recovery status (the raw Elasticsearch indices.recovery payload, keyed by index name).

Authentication: Required

Get index statusGET /api/datasets/admin:inventory-levels/index/status

Response · 200

{
  "_links": {
    "self": {
      "href": "https://informer.example.com/api/datasets/admin%3Ainventory-levels/index/status"
    }
  },
  "docsex.acme.inventory-levels.1700000000000": {
    "shards": [
      {
        "id": 0,
        "type": "EMPTY_STORE",
        "stage": "DONE",
        "primary": true,
        "start_time_in_millis": 1700000000000,
        "stop_time_in_millis": 1700000000000,
        "total_time_in_millis": 20,
        "source": {},
        "target": {
          "id": "xKGDDOQJTjmfkyqdAgqqNg",
          "host": "172.20.0.2",
          "transport_address": "172.20.0.2:9300",
          "ip": "172.20.0.2",
          "name": "87d1ec9ac8ec"
        },
        "index": {
          "size": {
            "total_in_bytes": 0,
            "reused_in_bytes": 0,
            "recovered_in_bytes": 0,
            "recovered_from_snapshot_in_bytes": 0,
            "percent": "0.0%"
          },
          "files": {
            "total": 0,
            "reused": 0,
            "recovered": 0,
            "percent": "0.0%"
          },
          "total_time_in_millis": 20,
          "source_throttle_time_in_millis": 0,
          "target_throttle_time_in_millis": 0
        },
        "translog": {
          "recovered": 0,
          "total": 0,
          "percent": "100.0%",
          "total_on_start": 0,
          "total_time_in_millis": 20
        },
        "verify_index": {
          "check_index_time_in_millis": 0,
          "total_time_in_millis": 20
        }
      }
    ]
  }
}

Captured from the API examples test suite; ids and timestamps are normalized.

Each shard entry reports its recovery type, stage (DONE when fully recovered), and per-phase progress. This is operational detail; for simple "how big / how many" questions use the count and size endpoints below.

---

GET /api/datasets/{id}/index/count

Get the number of documents in the Dataset's index. The response body is the bare number.

Authentication: Required

Get index document countGET /api/datasets/admin:inventory-levels/index/count

Response · 200

3

Captured from the API examples test suite; ids and timestamps are normalized.

---

GET /api/datasets/{id}/index/size

Get the index size in bytes. The response body is the bare number.

Authentication: Required

Get index size in bytesGET /api/datasets/admin:inventory-levels/index/size

Response · 200

6292

Captured from the API examples test suite; ids and timestamps are normalized.

---

DELETE /api/es-index/{id}

Delete a search index by name (admin operation).

Authentication: Required

Permission: Superuser

Path Parameters:

Parameter	Type	Description
id	string	Full index name (URL-encode it; index names contain dots)

Delete a search index (superuser)DELETE /api/es-index/docsex.acme.temp-import-a.1700000000000

Deletes the index and clears esIndex/records/size on any Dataset that pointed at it; the Dataset row itself survives. 404 when no such index exists. Responds 200 with an empty body.

Response · 200 · no body

Captured from the API examples test suite; ids and timestamps are normalized.

Behavior:

• Deletes the index, including flow-step scratch indices that have not expired yet
• Any Dataset pointing at the index has its esIndex, esType, records, and size cleared; the Dataset row itself survives and can be refreshed again
• 404 when no index by that name exists; 200 with an empty body on success

Administrative Operation

This bypasses Dataset-level permissions and deletes index data directly. Deleted indices cannot be recovered (refresh rebuilds Datasource-backed Datasets, but inline/upload data is gone).

---

POST /api/vacuum-es-indices

Delete multiple search indices in one call (admin operation).

Authentication: Required

Permission: Superuser

Request Body:

Field	Type	Required	Description
indices	array	Yes	Index names to delete

Delete several search indexes (superuser)POST /api/vacuum-es-indices

Same effect as the single delete, for a batch of index names. Responds 200 with an empty body.

Request body

{
  "indices": [
    "docsex.acme.temp-import-b.1700000000000"
  ]
}

Response · 200 · no body

Captured from the API examples test suite; ids and timestamps are normalized.

Behavior:

• Same per-index effect as DELETE /api/es-index/{id}, applied sequentially
• 404 if any named index does not exist (indices earlier in the list are already deleted at that point)
• 200 with an empty body on success

Use Case:

Batch cleanup of orphaned indices identified by the vacuum report.