[NIFI-10687] PutElasticsearchRecord/Json incorrectly try to use empty String as document _id

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.18.0
• Fix Version/s: 1.19.0
• Assignee: Chris Sampson
• Reporter: Chris Sampson
• Created: 24/Oct/22 19:25
• Updated: 25/Oct/22 18:24
• Resolved: 25/Oct/22 18:24

Description

PutElasticsearchRecord and PutElasticsearchJson will try to pass an empty/blank String as the _id of the document to be created in Elasticsearch if the processor is, for example, configured to use an attribute from the FlowFile or a field from the Record that doesn't exist.

To re-create:

• set the ID field to point at a non-existent Field using Expression Language, e.g. ${notExist}
• input a FlowFile where the attribute/Record Path equates to a non-existent Field (i.e. returns an empty String, but not null)
• observe that the attempt to send a document to Elasticsearch uses the empty String rather than null

Setting the _id in NiFi to null in such circumstances would be preferable/expected as this would cause Elasticsearch to auto-generate an _id when indexing.

Other NiFi processor fields may also be affected, e.g. for @timestamp (others should be considered, but some fields should legitimately equate to an empty/blank String and cause an error - index, for example)

Issue Links

• relates to: NIFI-10303 GetElasticsearch processor invokes elasticsearch GET api without an identifier when flow file is null
• links to: GitHub Pull Request #6575