Deleting documents in Elasticsearch - Step-by-Step Execution

Concept Flow - Deleting documents

Start

↓

Send Delete Request

↓

Check Document Exists?

No→Return Not Found

Yes↓

Delete Document

↓

Return Success Response

↓

End

The flow shows sending a delete request, checking if the document exists, deleting it if found, and returning the result.

Execution Sample

Elasticsearch

DELETE /my_index/_doc/1

# Deletes document with ID 1 from 'my_index'

This command deletes the document with ID 1 from the index named 'my_index'.

Execution Table

Step	Action	Request Sent	Document Exists?	Result	Response
1	Send DELETE request to /my_index/_doc/1	DELETE /my_index/_doc/1	Check if document with ID 1 exists	Yes	Proceed to delete
2	Delete document with ID 1	N/A	N/A	Deleted	200 OK with result 'deleted'
3	Return response	N/A	N/A	Success	{"result":"deleted"}
4	End	N/A	N/A	N/A	N/A

💡 Document with ID 1 deleted successfully, operation ends.

Variable Tracker

Variable	Start	After Step 1	After Step 2	Final
document_exists	unknown	true	deleted	deleted
response	none	pending	{"result":"deleted"}	{"result":"deleted"}

Key Moments - 2 Insights

What happens if the document does not exist when the delete request is sent?

If the document does not exist, the response will indicate 'not_found' instead of 'deleted'. This is shown in the execution_table step 1 where document existence is checked.

Does the delete request remove the entire index?

No, the delete request targets a single document by its ID. The index remains intact. This is clear from the request path /my_index/_doc/1 which specifies one document.

Visual Quiz - 3 Questions

Look at the execution table, what is the response after the document is deleted?

A {"result":"deleted"}

B {"result":"not_found"}

C 404 Not Found

D 500 Internal Server Error

💡 Hint

Check the Response column at Step 3 in the execution_table.

At which step does the system confirm the document exists before deleting?

A Step 3

B Step 1

C Step 2

D Step 4

💡 Hint

Look at the Document Exists? column in the execution_table.

If the document did not exist, how would the response change?

A Response would be 200 OK with no body

B Response would be {"result":"deleted"}

C Response would be {"result":"not_found"}

D Response would be 500 Internal Server Error

💡 Hint

Refer to key_moments about document existence and response.

Concept Snapshot

Deleting documents in Elasticsearch:
- Use DELETE /index/_doc/document_id
- Checks if document exists
- If yes, deletes and returns {"result":"deleted"}
- If no, returns {"result":"not_found"}
- Does not delete the whole index

Full Transcript

This visual execution shows how deleting a document in Elasticsearch works step-by-step. First, a DELETE request is sent to the document's URL. The system checks if the document exists. If it does, the document is deleted and a success response with {"result":"deleted"} is returned. If the document does not exist, the response indicates {"result":"not_found"}. The index itself remains unchanged. Variables like document_exists and response update as the process runs. This helps beginners see exactly what happens behind the scenes when deleting documents.

Practice

1.

Which Elasticsearch API method is used to delete a single document by its ID?

easy

A. GET method with document ID

B. DELETE method with document ID

C. POST method with document ID

D. PUT method with document ID

Solution

1. Step 1: Understand document deletion in Elasticsearch

   Deleting a single document requires specifying its ID and using the DELETE HTTP method.

2. Step 2: Identify the correct API method

   The DELETE method with the document ID is the standard way to remove one document.

3. Final Answer:

   DELETE method with document ID -> Option B

4. Quick Check:

   Delete single document = DELETE with ID [OK]

Hint: Use DELETE with document ID to remove one document fast [OK]

Common Mistakes:

• Using GET or POST instead of DELETE
• Trying to delete without specifying document ID
• Confusing _delete_by_query with single document deletion

2.

Which of the following is the correct syntax to delete a document with ID 123 from index products using Elasticsearch REST API?

?

easy

A. GET /products/_delete/123

B. POST /products/_doc/123/_delete

C. DELETE /products/_doc/123

D. PUT /products/_doc/123

Solution

1. Step 1: Recall the REST API syntax for deleting a document

   The correct syntax uses DELETE method on the path /index/_doc/document_id.

2. Step 2: Match the syntax with the options

   DELETE /products/_doc/123 matches DELETE /products/_doc/123 which is correct.

3. Final Answer:

   DELETE /products/_doc/123 -> Option C

4. Quick Check:

   Delete syntax = DELETE /index/_doc/id [OK]

Hint: DELETE + /index/_doc/document_id is the right syntax [OK]

Common Mistakes:

• Using POST or GET instead of DELETE
• Wrong URL path like _delete or missing _doc
• Using PUT which is for creating/updating

3.

Given this Elasticsearch _delete_by_query request to delete documents where status is "inactive":

{
  "query": {
    "term": { "status": "inactive" }
  }
}

What will be the result if there are 3 documents matching this query?

medium

A. Deletes 3 documents and returns success count 3

B. Deletes 0 documents because term query is incorrect

C. Deletes all documents in the index

D. Returns an error due to missing document ID

Solution

1. Step 1: Understand _delete_by_query behavior

   It deletes all documents matching the query, here documents with status "inactive".

2. Step 2: Apply to given query and data

   Since 3 documents match, all 3 will be deleted and success count will be 3.

3. Final Answer:

   Deletes 3 documents and returns success count 3 -> Option A

4. Quick Check:

   _delete_by_query deletes matching docs = 3 [OK]

Hint: Delete by query removes all matching docs, not just one [OK]

Common Mistakes:

• Thinking _delete_by_query needs document ID
• Assuming it deletes all documents regardless of query
• Confusing term query syntax

4.

What is wrong with this Elasticsearch delete request?

DELETE /users/_doc

It aims to delete a document but fails. Why?

medium

A. Missing document ID in the URL

B. Using DELETE instead of POST

C. Index name is incorrect

D. Missing request body with query

Solution

1. Step 1: Check the delete request URL format

   Deleting a document requires the document ID after _doc in the URL.

2. Step 2: Identify the missing part

   The request lacks the document ID, so Elasticsearch cannot identify which document to delete.

3. Final Answer:

   Missing document ID in the URL -> Option A

4. Quick Check:

   Delete URL needs document ID [OK]

Hint: Always include document ID after _doc in DELETE URL [OK]

Common Mistakes:

• Omitting document ID in delete URL
• Confusing DELETE with POST for deletion
• Thinking request body is needed for single delete

5.

You want to delete all documents from the orders index where the amount field is less than 100. Which Elasticsearch request correctly achieves this?

hard

A. DELETE /orders/_delete_by_query { "query": { "term": { "amount": 100 } } }

B. DELETE /orders/_doc { "query": { "range": { "amount": { "lt": 100 } } } }

C. POST /orders/_doc/amount_lt_100/_delete

D. POST /orders/_delete_by_query { "query": { "range": { "amount": { "lt": 100 } } } }

Solution

1. Step 1: Identify the correct API for deleting multiple documents by condition

   _delete_by_query is used with POST to delete documents matching a query.

2. Step 2: Check the query for amount less than 100

   The range query with "lt": 100 correctly matches documents with amount less than 100.

3. Step 3: Verify the request method and path

   POST /orders/_delete_by_query with the correct JSON body is the proper syntax.

4. Final Answer:

   POST /orders/_delete_by_query { "query": { "range": { "amount": { "lt": 100 } } } } -> Option D

5. Quick Check:

   Delete by query with range and POST = correct [OK]

Hint: Use POST _delete_by_query with range query for conditional deletes [OK]

Common Mistakes:

• Using DELETE method with request body (not supported)
• Using _doc endpoint instead of _delete_by_query
• Using term query instead of range for less than condition