fix(backend): speedup load_schema_from_db by removing pagination

Signed-off-by: Fatih Acar <fatih@opsmill.com>

backend/infrahub/core/schema/manager.py

@@ -2,7 +2,9 @@
 
 from typing import TYPE_CHECKING, Any, Optional, Union
 
-from infrahub import lock
+from opentelemetry import trace
+
+from infrahub import config, lock
 from infrahub.core.manager import NodeManager
 from infrahub.core.models import (
     HashableModelDiff,
@@ -607,6 +609,7 @@ async def load_schema(
         self.set_schema_branch(name=branch.name, schema=branch_schema)
         return branch_schema
 
+    @trace.get_tracer(__name__).start_as_current_span("load_schema_from_db")
     async def load_schema_from_db(
         self,
         db: InfrahubDatabase,
@@ -661,6 +664,9 @@ async def load_schema_from_db(
                 if removed_node in schema.node_names:
                     schema.delete(name=removed_node)
 
+        old_query_size_limit = config.SETTINGS.database.query_size_limit
+        config.SETTINGS.database.query_size_limit = 100000  # we grab 15% performance due to pagination in SchemaNodes when doing _enrich_node_dicts_with_relationships
+
         if not has_filters or filters["generics"]:
             generic_schema = self.get(name="SchemaGeneric", branch=branch)
             for schema_node in await self.query(
@@ -688,6 +694,8 @@ async def load_schema_from_db(
                     schema=await self.convert_node_schema_to_schema(schema_node=schema_node, db=db),
                 )
 
+        config.SETTINGS.database.query_size_limit = old_query_size_limit
+
         schema.process(validate_schema=validate_schema)
 
         return schema