[BUG] Requests are send to wrong port despite settings

[BlackRoseonesixone]

When attempting to vectorize the current chat, requests are send to port 8080 rather than the configured port 7850.
Alternative endpoints are configured both in ST and VectHare. ST is 1.14.0 and both VectHare and Similharity are at latest versions. OS is Windows 11 24H2

---

[LanceDB] Generating embeddings for 5 items
[similharity] chunks/insert error: FetchError: request to http://localhost:8080/v1/embeddings failed, reason:
    at ClientRequest.<anonymous> (file:///L:/SillyTavern/node_modules/node-fetch/src/index.js:108:11)
    at ClientRequest.emit (node:events:536:35)
    at emitErrorEvent (node:_http_client:104:11)
    at Socket.socketErrorListener (node:_http_client:518:5)
    at Socket.emit (node:events:524:28)
    at emitErrorNT (node:internal/streams/destroy:170:8)
    at emitErrorCloseNT (node:internal/streams/destroy:129:3)
    at process.processTicksAndRejections (node:internal/process/task_queues:90:21) {
  type: 'system',
  errno: 'ECONNREFUSED',
  code: 'ECONNREFUSED',
  erroredSysCall: undefined
}

---

---

╔══════════════════════════════════════════════════════════════╗
║              VECTHARE DIAGNOSTICS REPORT                      ║
╚══════════════════════════════════════════════════════════════╝

📅 Generated: 15/12/2025, 00:36:51
📦 Extension Version: 2.0.1-alpha
🔌 Plugin Version: 3.1.0

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
                      CURRENT SETTINGS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

  Backend:           lancedb
  Embedding Source:  llamacpp
  Model:             n/a (provider handles it)
  Provider URL:      http://127.0.0.1:7850
  Chunk Size:        500 chars (adaptive only)
  Score Threshold:   0.25
  Query Depth:       2
  Chat Auto-Sync:    disabled

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
                         SUMMARY
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

  ✓ Passed:    55
  ⚠ Warnings:   3
  ✗ Failed:    11
  ─────────────────
  Total:       69

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🔧 INFRASTRUCTURE  (✓8 ⚠0 ✗0)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

  ✓ ST Vectra (Standard)         Standard file-based vector storage ready
  ✓ ST Vector Endpoints          All 5 endpoints: list ✓, query ✓, insert ✓, delete ✓, purge ✓
  ✓ VectHare Plugin              v3.1.0 - Features: core
  ✓ Plugin API Endpoints         6 endpoints: health✓, collections✓, sources✓, chunks/list✓, chunks/query✓, backend/health✓
  ✓ LanceDB (Scalable)           LanceDB ready - disk-based vector storage active
  ○ Qdrant (Production)          Not selected (using: lancedb)
  ○ Qdrant Dimensions            Not using Qdrant backend
  ✓ Embedding Provider           Using llama.cpp
  ✓ API URL                      Custom: http://127.0.0.1:7850
  ✓ Provider Connectivity        Provider llama.cpp is recognized

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚙️  CONFIGURATION  (✓16 ⚠1 ✗1)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

  ✓ RAG Query Status             Other collections: 4 enabled
  ✓ Chunk Size                   500 characters
  ✓ Score Threshold              0.25
  ✓ Insert/Query Counts          Insert: 3, Query: 2
  ⚠ Chat Vectors                 Current chat has no vectorized chunks
  ✓ Temporal Decay               Per-collection decay active. 0 chunk(s) temporally blind.
  ✓ Temporally Blind Chunks      No chunks marked as temporally blind
  ✓ Conditional Activation       Module loaded (28 emotions, 5 gen types)
  ✗ Collection ID Format         Collection ID missing 'vh' prefix: vecthare_chat_assistant_b489fc69-cff8-46cc-96a0-87ec6c0204ba
  ✓ Hash Collision Rate          No chunks to analyze
  ✓ Chat Metadata Integrity      UUID: b489fc69...
  ✓ Condition Rules              No chunks with conditions
  ✓ Collection Registry          4 collection(s) registered, 4 enabled for querying
  ✓ Chunk Groups Module          Module loaded and functional
  ✓ Chunk Groups Validity        No groups configured
  ✓ Group Member Integrity       No group members to check
  ✓ Prompt Context               Global: context+xml
  ✓ PNG Export Capability        Full PNG export support (native deflate-raw)

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
👁️  VISUALIZER  (✓3 ⚠0 ✗6)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

  ✓ Visualizer Settings          Backend: lancedb, Source: llamacpp
  ✓ Hash Generation              Working correctly (test hash: 4568356664438633)
  ✓ Metadata Operations          Save, read, and delete all working
  ✗ Vector Insert                Insert failed: [LanceDB] Failed to insert 1 vectors into vh:test:visualizer_1765755407395: 500 Internal Server Error - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
  ✗ Vector Delete                Delete failed: [LanceDB] Failed to insert 1 vectors into vh:test:visualizer_1765755407458: 500 Internal Server Error - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
  ✗ Re-Vectorization             Workflow failed: [LanceDB] Failed to insert 1 vectors into vh:test:visualizer_1765755407542: 500 Internal Server Error - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
  ✗ Summary Vector Create        Failed: [LanceDB] Failed to insert 1 vectors into vh:test:visualizer_1765755407609: 500 Internal Server Error - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
  ✗ Summary Vector Delete        Failed: [LanceDB] Failed to insert 2 vectors into vh:test:visualizer_1765755407731: 500 Internal Server Error - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
  ✗ Backend Responsiveness       Backend error: [LanceDB] Failed to insert 1 vectors into vh:test:visualizer_1765755407819: 500 Internal Server Error - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🧪 PRODUCTION TESTS  (✓28 ⚠2 ✗4)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

  ✓ [PROD] Embedding Generation  Successfully generated test embedding (2560 dimensions)
  ✗ [PROD] Vector Storage        Failed to store vector: 500 - {"error":"lance error: LanceError(IO): Generic LocalFileSystem error: Unable to copy file from L:\\SillyTavern\\data\\user\\vectors\\lancedb\\llamacpp\\vh_test_storage_1765755410231.lance\\_versions\\.tmp_1.manifest_4dfdf0ac-3ea9-4696-83a4-1b31ab073616 to L:\\SillyTavern\\data\\user\\vectors\\lancedb\\llamacpp\\vh_test_storage_1765755410231.lance\\_versions\\1.manifest: Incorrect function. (os error 1), C:\\Users\\runneradmin\\.cargo\\registry\\src\\index.crates.io-6f17d22bba15001f\\lance-table-0.12.1\\src\\io\\commit.rs:692:54"}
  ⚠ [PROD] Vector Retrieval      No vectors in current chat to test retrieval
  ✓ [PROD] Vector Dimensions     No vectors stored yet - dimensions will be set on first vectorization
  ✗ [PROD] Temporal Decay        Decay not reducing scores (check formula)
  ✗ [PROD] Temporally Blind Chunks Blind chunk score was modified
  ⚠ [PROD] Chunk-Server Sync     11 orphaned local entries (vectors deleted from server)
  ✓ [PROD] Duplicate Hash Check  No vectors in collection
  ✗ [PROD] Plugin Embedding Gen  Plugin failed to generate embedding: 500 - {"error":"lance error: LanceError(IO): Generic LocalFileSystem error: Unable to copy file from L:\\SillyTavern\\data\\user\\vectors\\lancedb\\llamacpp\\vh_test_embed_1765755411311.lance\\_versions\\.tmp_1.manifest_e9ebbb20-6fe1-4712-b738-9bf5cf3e10e1 to L:\\SillyTavern\\data\\user\\vectors\\lancedb\\llamacpp\\vh_test_embed_1765755411311.lance\\_versions\\1.manifest: Incorrect function. (os error 1), C:\\Users\\runneradmin\\.cargo\\registry\\src\\index.crates.io-6f17d22bba15001f\\lance-table-0.12.1\\src\\io\\commit.rs:692:54"}
  ✓ Condition: Context Build     Built context with 5 messages
  ✓ Condition: No Conditions     Chunks without conditions pass
  ✓ Condition: Pattern Match     Found "happy"/"excited" in chat
  ✓ Condition: Pattern Regex     Regex /adventur\w*/i matched
  ✓ Condition: Speaker           Found User in speakers
  ✓ Condition: Message Count     Chat has >= 3 messages
  ✓ Condition: Character Present user is present
  ✓ Condition: Generation Type   Type is "normal"
  ✓ Condition: Group Chat        Not a group chat
  ✓ Condition: Lorebook Active   "adventure" entry active
  ✓ Condition: Random Chance     100% probability passed
  ✓ Condition: Time of Day       Within 00:00-23:59
  ✓ Condition: Emotion           Detected joy/excitement
  ✓ Condition: Swipe Count       Swipe count >= 1
  ✓ Condition: Score Threshold   0.75 >= 0.5 threshold
  ✓ Condition: Recency           Chunk is >= 50 msgs old
  ✓ Condition: Frequency         Under max activations
  ✓ Condition: AND Logic         Both conditions passed
  ✓ Condition: OR Logic          One condition passed
  ✓ Condition: Negation          NOT isGroupChat worked
  ✓ Condition: Soft Links        Soft-linked chunk boosted
  ✓ Condition: Hard Links        Missing hard link detected
  ✓ Condition: Filter Chunks     Filtered to 2 chunks (expected 2)
  ✓ Condition: Validation Valid  Valid rule accepted
  ✓ Condition: Validation Invalid Invalid rule rejected

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
                      END OF REPORT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

---

[eapeyton205]

+1 to this. Endpoints are set to BB server correctly in both the llamacpp connection settings and in vecthare itself, but it appears to be trying to use the default llamacpp port anyway. Elsewhere in the error log it shows the correct 8008 port (under API URL, but something is going wonky here.)

[eapeyton205]

╔══════════════════════════════════════════════════════════════╗
║ VECTHARE DIAGNOSTICS REPORT ║
╚══════════════════════════════════════════════════════════════╝

📅 Generated: 12/19/2025, 6:10:10 PM
📦 Extension Version: 2.0.3-alpha
🔌 Plugin Version: 3.1.0

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
CURRENT SETTINGS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Backend: standard
Embedding Source: llamacpp
Model: n/a (provider handles it)
Provider URL: http://localhost:8008
Chunk Size: 500 chars (adaptive only)
Score Threshold: 0.25
Query Depth: 2
Chat Auto-Sync: disabled

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
SUMMARY
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

✓ Passed: 55
⚠ Warnings: 2
✗ Failed: 10
─────────────────
Total: 67

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🔧 INFRASTRUCTURE (✓7 ⚠0 ✗0)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

✓ ST Vectra (Standard) Standard file-based vector storage ready
✓ ST Vector Endpoints All 5 endpoints: list ✓, query ✓, insert ✓, delete ✓, purge ✓
✓ VectHare Plugin v3.1.0 - Features: core
✓ Plugin API Endpoints 6 endpoints: health✓, collections✓, sources✓, chunks/list✓, chunks/query✓, backend/health✓
○ LanceDB (Scalable) Not selected (using: standard)
○ Qdrant (Production) Not selected (using: standard)
○ Qdrant Dimensions Not using Qdrant backend
✓ Embedding Provider Using llama.cpp
✓ API URL http://127.0.0.1:8008/
✓ Provider Connectivity Provider llama.cpp is recognized

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚙️ CONFIGURATION (✓17 ⚠0 ✗1)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

✓ RAG Query Status Other collections: 19 enabled
✓ Chunk Size 500 characters
✓ Score Threshold 0.25
✓ Insert/Query Counts Insert: 3, Query: 2
✓ Chat Vectors 404 vectorized chunks
✓ Temporal Decay Per-collection decay active. 0 chunk(s) temporally blind.
✓ Temporally Blind Chunks No chunks marked as temporally blind
✓ Conditional Activation Module loaded (28 emotions, 5 gen types)
✗ Collection ID Format Collection ID missing 'vh' prefix: vecthare_chat_augustus_59c11e05-1d13-46c0-a14f-5f1cb752f69f
✓ Hash Collision Rate 404 chunks, all unique hashes
✓ Chat Metadata Integrity UUID: 59c11e05...
✓ Condition Rules No chunks with conditions
✓ Collection Registry 19 collection(s) registered, 19 enabled for querying
✓ Chunk Groups Module Module loaded and functional
✓ Chunk Groups Validity No groups configured
✓ Group Member Integrity No group members to check
✓ Prompt Context Global: context+xml
✓ PNG Export Capability Full PNG export support (native deflate-raw)

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
👁️ VISUALIZER (✓3 ⚠0 ✗6)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

✓ Visualizer Settings Backend: standard, Source: llamacpp
✓ Hash Generation Working correctly (test hash: 4568356664438633)
✓ Metadata Operations Save, read, and delete all working
✗ Vector Insert Insert failed: Failed to insert vectors: 500 - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
✗ Vector Delete Delete failed: Failed to insert vectors: 500 - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
✗ Re-Vectorization Workflow failed: Failed to insert vectors: 500 - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
✗ Summary Vector Create Failed: Failed to insert vectors: 500 - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
✗ Summary Vector Delete Failed: Failed to insert vectors: 500 - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
✗ Backend Responsiveness Backend error: Failed to insert vectors: 500 - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🧪 PRODUCTION TESTS (✓28 ⚠2 ✗3)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

✗ [PROD] Embedding Generation Failed to generate embedding: 500 Internal Server Error - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
✗ [PROD] Vector Storage Failed to store vector: 500 - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
✗ [PROD] Vector Retrieval Query failed: 500
⚠ [PROD] Vector Dimensions Could not generate test embedding: 500 - {"error":"request to http://localhost:8080/v1/embeddings failed, reason: "}
✓ [PROD] Temporal Decay Decay working (0.85 -> 0.425 at age 50 for chat)
✓ [PROD] Temporally Blind Chunks Blind: 0.90 (immune), Normal: 0.27 (decayed)
⚠ [PROD] Chunk-Server Sync 18 orphaned local entries (vectors deleted from server)
✓ [PROD] Duplicate Hash Check 404 unique vectors, no duplicates
○ [PROD] Plugin Embedding Gen Standard backend uses native ST vectors
✓ Condition: Context Build Built context with 5 messages
✓ Condition: No Conditions Chunks without conditions pass
✓ Condition: Pattern Match Found "happy"/"excited" in chat
✓ Condition: Pattern Regex Regex /adventur\w*/i matched
✓ Condition: Speaker Found User in speakers
✓ Condition: Message Count Chat has >= 3 messages
✓ Condition: Character Present Luna is present
✓ Condition: Generation Type Type is "normal"
✓ Condition: Group Chat Not a group chat
✓ Condition: Lorebook Active "adventure" entry active
✓ Condition: Random Chance 100% probability passed
✓ Condition: Time of Day Within 00:00-23:59
✓ Condition: Emotion Detected joy/excitement
✓ Condition: Swipe Count Swipe count >= 1
✓ Condition: Score Threshold 0.75 >= 0.5 threshold
✓ Condition: Recency Chunk is >= 50 msgs old
✓ Condition: Frequency Under max activations
✓ Condition: AND Logic Both conditions passed
✓ Condition: OR Logic One condition passed
✓ Condition: Negation NOT isGroupChat worked
✓ Condition: Soft Links Soft-linked chunk boosted
✓ Condition: Hard Links Missing hard link detected
✓ Condition: Filter Chunks Filtered to 2 chunks (expected 2)
✓ Condition: Validation Valid Valid rule accepted
✓ Condition: Validation Invalid Invalid rule rejected

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
END OF REPORT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━