refactor(embedding-cache): rename index cache to embedding cache

The cached payload is the indexing pipeline's embeddings (markdown is
chunked then embedded), so "embedding cache" names the expensive output
directly and removes the "index" ambiguity (DB index vs vector index vs
indexing phase). Renames the service, settings, eligibility, eviction
task, metrics, config flags (INDEX_CACHE_* -> EMBEDDING_CACHE_*), object
prefix, and the table (index_cache_embedding_sets -> embedding_cache_sets)
with its constraint and indexes. Migration 161 renamed accordingly.

surfsense_backend/app/observability/metrics.py

@@ -306,18 +306,18 @@ def _etl_cache_evictions():
 
 
 @lru_cache(maxsize=1)
-def _index_cache_lookups():
+def _embedding_cache_lookups():
     return _get_meter().create_counter(
-        "surfsense.index.cache.lookups",
-        description="Count of index (chunk+embedding) cache lookups by outcome (hit/miss).",
+        "surfsense.embedding.cache.lookups",
+        description="Count of embedding (chunk+embedding) cache lookups by outcome (hit/miss).",
     )
 
 
 @lru_cache(maxsize=1)
-def _index_cache_evictions():
+def _embedding_cache_evictions():
     return _get_meter().create_counter(
-        "surfsense.index.cache.evictions",
-        description="Count of index cache entries evicted, by phase.",
+        "surfsense.embedding.cache.evictions",
+        description="Count of embedding cache entries evicted, by phase.",
     )
 
 
@@ -724,12 +724,12 @@ def record_etl_cache_eviction(count: int, *, phase: str) -> None:
     _add(_etl_cache_evictions(), count, {"phase": phase})
 
 
-def record_index_cache_lookup(
+def record_embedding_cache_lookup(
     *, embedding_model: str | None, chunker_kind: str | None, outcome: str
 ) -> None:
-    """Record an index-cache lookup. ``outcome`` is ``hit`` or ``miss``."""
+    """Record an embedding-cache lookup. ``outcome`` is ``hit`` or ``miss``."""
     _add(
-        _index_cache_lookups(),
+        _embedding_cache_lookups(),
         1,
         {
             "embedding.model": embedding_model or "unknown",
@@ -739,11 +739,11 @@ def record_index_cache_lookup(
     )
 
 
-def record_index_cache_eviction(count: int, *, phase: str) -> None:
+def record_embedding_cache_eviction(count: int, *, phase: str) -> None:
     """Record evicted entries. ``phase`` is ``ttl`` or ``size``."""
     if count <= 0:
         return
-    _add(_index_cache_evictions(), count, {"phase": phase})
+    _add(_embedding_cache_evictions(), count, {"phase": phase})
 
 
 def record_celery_heartbeat_refresh(*, heartbeat_type: str) -> None:
@@ -942,12 +942,12 @@ __all__ = [
     "record_compaction_run",
     "record_connector_sync_duration",
     "record_connector_sync_outcome",
+    "record_embedding_cache_eviction",
+    "record_embedding_cache_lookup",
     "record_etl_cache_eviction",
     "record_etl_cache_lookup",
     "record_etl_extract_duration",
     "record_etl_extract_outcome",
-    "record_index_cache_eviction",
-    "record_index_cache_lookup",
     "record_indexing_document_duration",
     "record_indexing_document_outcome",
     "record_interrupt",