Service Layer

Relevant source files

• docs/docs/API-Reference/curl-examples/api-reference-api-examples/result-get-configuration.json
• docs/docs/Develop/api-keys-and-authentication.mdx
• docs/docs/Develop/memory.mdx
• docs/docs/Flows/webhook.mdx
• src/backend/base/langflow/api/v1/variable.py
• src/backend/base/langflow/services/__init__.py
• src/backend/base/langflow/services/cache/base.py
• src/backend/base/langflow/services/cache/factory.py
• src/backend/base/langflow/services/cache/service.py
• src/backend/base/langflow/services/cache/utils.py
• src/backend/base/langflow/services/chat/service.py
• src/backend/base/langflow/services/factory.py
• src/backend/base/langflow/services/flow_events/factory.py
• src/backend/base/langflow/services/flow_events/service.py
• src/backend/base/langflow/services/session/service.py
• src/backend/base/langflow/services/store/factory.py
• src/backend/base/langflow/services/telemetry/__init__.py
• src/backend/base/langflow/services/telemetry/factory.py
• src/backend/base/langflow/services/telemetry/opentelemetry.py
• src/backend/base/langflow/services/variable/__init__.py
• src/backend/base/langflow/services/variable/base.py
• src/backend/base/langflow/services/variable/factory.py
• src/backend/base/langflow/services/variable/kubernetes.py
• src/backend/base/langflow/services/variable/service.py
• src/backend/tests/unit/api/v1/test_mcp_reverse_proxy.py
• src/backend/tests/unit/api/v1/test_variable.py
• src/backend/tests/unit/services/test_cache_factory.py
• src/backend/tests/unit/services/variable/test_service.py
• src/backend/tests/unit/test_auth_jwt_algorithms.py
• src/backend/tests/unit/test_flow_events_service.py
• src/backend/tests/unit/test_security_cors.py
• src/backend/tests/unit/test_webhook.py
• src/backend/tests/unit/test_webhook_sse_regression.py
• src/lfx/src/lfx/services/manager.py
• src/lfx/src/lfx/services/settings/auth.py
• src/lfx/src/lfx/services/settings/base.py
• src/lfx/tests/unit/services/settings/test_runtime_port.py

This page documents the service layer of the Langflow backend: the set of long-lived singleton services that underpin the application, how they are registered, initialized, and torn down, and the dependency injection mechanism used to access them from API endpoints and other modules.

For the configuration values that services read at startup, see Backend Settings Service. For the database schema and ORM models that services operate on, see Database Layer. For how setup_app() and the lifespan hook integrate with this layer, see Application Lifecycle.

---

Design: Service Registry Pattern

All application-level singletons are managed through a central service registry. Services are identified by values in the ServiceType enum (langflow.services.schema.ServiceType) and retrieved via helper functions defined in langflow.services.deps.

This pattern means:

• Services are instantiated once at application startup.
• Code that needs a service calls a dependency function, not a constructor.
• FastAPI endpoints use Depends(get_db_service) etc.; non-FastAPI code calls the same functions directly.

Core modules:

Module	Role
langflow.services.schema	ServiceType enum — the canonical name for every service src/backend/base/langflow/services/schema.py52
lfx.services.manager	ServiceManager class — manages the lifecycle and registration of services src/lfx/src/lfx/services/manager.py10-15
langflow.services.deps	get_service(), get_db_service(), get_auth_service(), get_variable_service(), etc. src/backend/base/langflow/services/deps.py45-50
langflow.services.utils	initialize_services(), initialize_settings_service(), teardown_services() src/backend/base/langflow/services/utils.py132-140

Sources: src/backend/base/langflow/services/utils.py132-140 src/backend/base/langflow/services/deps.py45-52 src/backend/base/langflow/services/schema.py52 src/lfx/src/lfx/services/manager.py10-15

---

Service Inventory

Services registered at startup (managed via ServiceManager):

Service	ServiceType name	Primary file	Purpose
Settings	SETTINGS_SERVICE	lfx.services.settings	Pydantic settings model; loaded from env/yaml before all others src/lfx/src/lfx/services/settings/base.py89-104
Database	DATABASE_SERVICE	langflow.services.database.service.DatabaseService	Async SQLAlchemy engine, session factory, Alembic migrations src/backend/base/langflow/services/database/service.py90-126
Auth	AUTH_SERVICE	langflow.services.auth	JWT signing, password hashing, API key validation src/backend/base/langflow/services/deps.py52
Cache	CACHE_SERVICE	langflow.services.cache	In-process or external async cache for flows and components src/backend/base/langflow/services/cache/service.py22-38
Storage	STORAGE_SERVICE	langflow.services.storage.service.StorageService	Local or S3-backed file storage src/backend/base/langflow/initial_setup/setup.py54
Variable	VARIABLE_SERVICE	langflow.services.variable.service.DatabaseVariableService	Global variable (credential) store per user src/backend/base/langflow/services/variable/service.py24-25
Telemetry	TELEMETRY_SERVICE	langflow.services.telemetry.opentelemetry.OpenTelemetry	OpenTelemetry metrics (Prometheus) and background telemetry src/backend/base/langflow/services/telemetry/opentelemetry.py108-115
Event Manager	EVENT_MANAGER	langflow.services.event_manager.WebhookEventManager	Manages webhooks and internal event propagation src/backend/tests/unit/test_webhook.py7

Sources: src/backend/base/langflow/services/utils.py132-165 src/backend/base/langflow/services/variable/service.py24-25 src/backend/base/langflow/services/telemetry/opentelemetry.py108-115 src/lfx/src/lfx/services/settings/base.py89-104

---

Initialization Sequence

Title: Service initialization flow inside get_lifespan and the startup lifespan context

The key ordering constraint is that initialize_settings_service() is called before the lifespan context is entered. This ensures get_settings_service() is available for early configuration src/backend/base/langflow/main.py151-152

Sources: src/backend/base/langflow/main.py150-176 src/backend/base/langflow/services/utils.py132-165 src/backend/base/langflow/services/cache/factory.py15-38

---

Dependency Injection

Title: Mapping service getters to backend code entities

Sources: src/backend/base/langflow/services/deps.py45-56 src/backend/base/langflow/services/variable/service.py24-25 src/backend/base/langflow/services/cache/service.py22

---

CacheService — Detailed

The Cache Service is initialized via CacheServiceFactory src/backend/base/langflow/services/cache/factory.py15-17 It supports multiple backends based on settings.cache_type src/backend/base/langflow/services/cache/factory.py24-38

Backend	Class	Characteristics
memory	ThreadingInMemoryCache	Thread-safe LRU using threading.RLock and OrderedDict src/backend/base/langflow/services/cache/service.py22-56
redis	RedisCache	External storage using Redis, supporting expiration src/backend/base/langflow/services/cache/service.py178-199
async	AsyncInMemoryCache	Non-blocking in-memory cache for async contexts src/backend/base/langflow/services/cache/factory.py36-37

The cache uses pickle or dill for serialization to mimic external store behavior even when running in-memory src/backend/base/langflow/services/cache/service.py78

Sources: src/backend/base/langflow/services/cache/service.py22-199 src/backend/base/langflow/services/cache/factory.py15-38

---

VariableService — Detailed

The VariableService manages user-defined variables and credentials. It has two primary implementations: DatabaseVariableService (SQL-backed) and KubernetesSecretService (K8s-backed) src/backend/base/langflow/services/variable/service.py24 src/backend/base/langflow/services/variable/kubernetes.py25

Key Features:

• Environment Initialization: On user initialization, it can sync environment variables (e.g., OPENAI_API_KEY) into the database if store_environment_variables is enabled src/backend/base/langflow/services/variable/service.py28-32
• Credential Protection: Variables of type CREDENTIAL_TYPE are encrypted using encrypt_api_key src/backend/base/langflow/services/variable/kubernetes.py168
• Validation: When creating model provider variables, it uses validate_model_provider_key to verify the key works before saving src/backend/base/langflow/api/v1/variable.py130-132
• Generic Types: Variables of type GENERIC_TYPE (like model lists) are stored as plaintext and are not encrypted src/backend/tests/unit/services/variable/test_service.py183-192

Sources: src/backend/base/langflow/services/variable/service.py24-140 src/backend/base/langflow/services/variable/kubernetes.py25-171 src/backend/base/langflow/api/v1/variable.py102-143

---

OpenTelemetry — Telemetry and Metrics

The telemetry service uses a thread-safe singleton pattern to manage OpenTelemetry metrics src/backend/base/langflow/services/telemetry/opentelemetry.py108-112

• Prometheus Integration: Uses PrometheusMetricReader to export metrics src/backend/base/langflow/services/telemetry/opentelemetry.py8
• Metric Types: Supports COUNTER, OBSERVABLE_GAUGE, HISTOGRAM, and UP_DOWN_COUNTER src/backend/base/langflow/services/telemetry/opentelemetry.py27-31
• Default Metrics: Automatically tracks file_uploads (gauge) and num_files_uploaded (counter) src/backend/base/langflow/services/telemetry/opentelemetry.py129-142

Sources: src/backend/base/langflow/services/telemetry/opentelemetry.py1-180

---

Settings Service

The Settings class is a composed Pydantic model that aggregates settings from multiple groups (Database, Cache, Security, etc.) src/lfx/src/lfx/services/settings/base.py89-104

• Env Prefix: Loads variables prefixed with LANGFLOW_ src/lfx/src/lfx/services/settings/base.py111
• YAML Support: Can update settings from a YAML file using update_from_yaml src/lfx/src/lfx/services/settings/base.py113-116
• Dynamic Updates: Supports runtime updates via update_settings, handling list-based environment variables (comma-separated) src/lfx/src/lfx/services/settings/base.py118-138

Sources: src/lfx/src/lfx/services/settings/base.py1-191