fix: added number_of_replicas as 0 to not create unused nodes in OpenSearch

[lucaseduoli]

This pull request focuses on optimizing OpenSearch index configurations by reducing the number of replicas to zero, which can improve indexing speed and reduce resource usage in development or single-node environments. The changes affect both the knowledge filters index and dynamically created embedding indices.

OpenSearch index configuration improvements:

• Set "number_of_replicas": 0 in the knowledge filters index settings in src/main.py to ensure no replicas are created for this index.
• Set "number_of_replicas": 0 (and remove the previous default of 1) for dynamically created embedding indices in src/utils/embeddings.py, further optimizing resource usage.

[mpawlow]

Functional Test

mpawlow@OpenRAG1:~/git/openrag$ make health
Health check:
Frontend:   Healthy
Backend:    Healthy
Langflow:   Healthy
OpenSearch: green
Docling:    Healthy

{
   "cluster_name":"docker-cluster",
   "status":"green",
   "timed_out":false,
   "number_of_nodes":1,
   "number_of_data_nodes":1,
   "discovered_master":true,
   "discovered_cluster_manager":true,
   "active_primary_shards":8,
   "active_shards":8,
   "relocating_shards":0,
   "initializing_shards":0,
   "unassigned_shards":0,
   "delayed_unassigned_shards":0,
   "number_of_pending_tasks":0,
   "number_of_in_flight_fetch":0,
   "task_max_waiting_in_queue_millis":0,
   "active_shards_percent_as_number":100.0
}

• ✅ Successfully sanity ingestion of documents from folder

[mpawlow]

@lucaseduoli

Code Review 1

• See PR comments: (a) to (d)

[mpawlow]

Code Review 2

• ✅ LGTM / Approved
• See: PR comment (2a)

[mpawlow]

Functional Test 2

• ✅ Successfully sanity ingestion of documents from folder