Indexing stats total include replicas

[andreidan]

Elasticsearch Version

8.14

Installed Plugins

No response

Java Version

bundled

OS Version

Darwin

Problem Description

Issuing a GET {index_name}/_stats/indexing request returns the primaries
result and the total result.
Some of the total metrics (like write_load) are averages and the values are being
misleading as they take into account the replicas values.
The indexing metrics in particular are mostly interesting off the primary shards.

This proposes we ignore replica indexing metrics in the total categore to avoid
outputs where an index with one primary and one replica reports some total metrics
as lower than the individual primary shard:

"indices": {
    ".ds-traces-apm-default-2024.03.26-000300": {
      "uuid": "o8gTCw_rQrmceYLSdJV_Pg",
      "health": "green",
      "status": "open",
      "primaries": {
        "indexing": {
          "index_total": 28735924,
          "index_time_in_millis": 4342901,
          "index_current": 1,
          "index_failed": 0,
          "delete_total": 0,
          "delete_time_in_millis": 0,
          "delete_current": 0,
          "noop_update_total": 0,
          "is_throttled": false,
          "throttle_time_in_millis": 0,
          "write_load": 1.7805114664625246
        }
      },
      "total": {
        "indexing": {
          "index_total": 28735924,
          "index_time_in_millis": 4342901,
          "index_current": 1,
          "index_failed": 0,
          "delete_total": 0,
          "delete_time_in_millis": 0,
          "delete_current": 0,
          "noop_update_total": 0,
          "is_throttled": false,
          "throttle_time_in_millis": 0,
          "write_load": 0.8903026091483134
        }
      }
    }

Steps to Reproduce

Create index/data stream
Ingest data
GET {index_name}/_stats/indexing

Logs (if relevant)

No response

[elasticsearchmachine]

Pinging @elastic/es-data-management (Team:Data Management)