add sentry (#2786)

* add sentry

* nit

* nit

* add requirement to ee

* try to ensure sentry is installed in integration tests

Author: pablonyx

backend/danswer/background/celery/celery_app.py

@@ -4,6 +4,7 @@
 from typing import Any
 
 import redis
+import sentry_sdk
 from celery import bootsteps  # type: ignore
 from celery import Celery
 from celery import current_task
@@ -16,6 +17,7 @@
 from celery.signals import worker_shutdown
 from celery.states import READY_STATES
 from celery.utils.log import get_task_logger
+from sentry_sdk.integrations.celery import CeleryIntegration
 
 from danswer.background.celery.celery_redis import RedisConnectorCredentialPair
 from danswer.background.celery.celery_redis import RedisConnectorDeletion
@@ -36,12 +38,24 @@
 from danswer.utils.logger import ColoredFormatter
 from danswer.utils.logger import PlainFormatter
 from danswer.utils.logger import setup_logger
+from shared_configs.configs import SENTRY_DSN
 
 logger = setup_logger()
 
 # use this within celery tasks to get celery task specific logging
 task_logger = get_task_logger(__name__)
 
+if SENTRY_DSN:
+    sentry_sdk.init(
+        dsn=SENTRY_DSN,
+        integrations=[CeleryIntegration()],
+        traces_sample_rate=0.5,
+    )
+    logger.info("Sentry initialized")
+else:
+    logger.debug("Sentry DSN not provided, skipping Sentry initialization")
+
+
 celery_app = Celery(__name__)
 celery_app.config_from_object(
     "danswer.background.celery.celeryconfig"

backend/danswer/main.py

@@ -5,6 +5,7 @@
 from typing import Any
 from typing import cast
 
+import sentry_sdk
 import uvicorn
 from fastapi import APIRouter
 from fastapi import FastAPI
@@ -15,6 +16,8 @@
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import JSONResponse
 from httpx_oauth.clients.google import GoogleOAuth2
+from sentry_sdk.integrations.fastapi import FastApiIntegration
+from sentry_sdk.integrations.starlette import StarletteIntegration
 from sqlalchemy.orm import Session
 
 from danswer import __version__
@@ -89,6 +92,7 @@
 from danswer.utils.variable_functionality import global_version
 from danswer.utils.variable_functionality import set_is_ee_based_on_env_variable
 from shared_configs.configs import CORS_ALLOWED_ORIGIN
+from shared_configs.configs import SENTRY_DSN
 
 logger = setup_logger()
 
@@ -201,6 +205,15 @@ def get_application() -> FastAPI:
     application = FastAPI(
         title="Danswer Backend", version=__version__, lifespan=lifespan
     )
+    if SENTRY_DSN:
+        sentry_sdk.init(
+            dsn=SENTRY_DSN,
+            integrations=[StarletteIntegration(), FastApiIntegration()],
+            traces_sample_rate=0.5,
+        )
+        logger.info("Sentry initialized")
+    else:
+        logger.debug("Sentry DSN not provided, skipping Sentry initialization")
 
     # Add the custom exception handler
     application.add_exception_handler(status.HTTP_400_BAD_REQUEST, log_http_error)

backend/model_server/main.py

@@ -4,9 +4,12 @@
 from contextlib import asynccontextmanager
 from pathlib import Path
 
+import sentry_sdk
 import torch
 import uvicorn
 from fastapi import FastAPI
+from sentry_sdk.integrations.fastapi import FastApiIntegration
+from sentry_sdk.integrations.starlette import StarletteIntegration
 from transformers import logging as transformer_logging  # type:ignore
 
 from danswer import __version__
@@ -19,6 +22,7 @@
 from shared_configs.configs import MIN_THREADS_ML_MODELS
 from shared_configs.configs import MODEL_SERVER_ALLOWED_HOST
 from shared_configs.configs import MODEL_SERVER_PORT
+from shared_configs.configs import SENTRY_DSN
 
 os.environ["TOKENIZERS_PARALLELISM"] = "false"
 os.environ["HF_HUB_DISABLE_TELEMETRY"] = "1"
@@ -81,6 +85,15 @@ def get_model_app() -> FastAPI:
     application = FastAPI(
         title="Danswer Model Server", version=__version__, lifespan=lifespan
     )
+    if SENTRY_DSN:
+        sentry_sdk.init(
+            dsn=SENTRY_DSN,
+            integrations=[StarletteIntegration(), FastApiIntegration()],
+            traces_sample_rate=0.5,
+        )
+        logger.info("Sentry initialized")
+    else:
+        logger.debug("Sentry DSN not provided, skipping Sentry initialization")
 
     application.include_router(management_router)
     application.include_router(encoders_router)

backend/requirements/default.txt

@@ -81,4 +81,5 @@ dropbox==11.36.2
 boto3-stubs[s3]==1.34.133
 ultimate_sitemap_parser==0.5
 stripe==10.12.0
-mistune==0.8.4
+mistune==0.8.4
+sentry-sdk==2.14.0

backend/requirements/ee.txt

@@ -1 +1 @@
-python3-saml==1.15.0
+python3-saml==1.15.0

backend/requirements/model_server.txt

@@ -12,4 +12,5 @@ torch==2.2.0
 transformers==4.39.2
 uvicorn==0.21.1
 voyageai==0.2.3
-litellm==1.49.5
+litellm==1.49.5
+sentry-sdk[fastapi,celery,starlette]==2.14.0

backend/shared_configs/configs.py

@@ -71,6 +71,10 @@
     os.environ.get("STRICT_CHUNK_TOKEN_LIMIT", "").lower() == "true"
 )
 
+# Set up Sentry integration (for error logging)
+SENTRY_DSN = os.environ.get("SENTRY_DSN")
+
+
 # Fields which should only be set on new search setting
 PRESERVED_SEARCH_FIELDS = [
     "id",

deployment/docker_compose/docker-compose.dev.yml

@@ -90,6 +90,9 @@ services:
       - LOG_POSTGRES_CONN_COUNTS=${LOG_POSTGRES_CONN_COUNTS:-}
       - CELERY_BROKER_POOL_LIMIT=${CELERY_BROKER_POOL_LIMIT:-}
 
+      # Analytics Configs
+      - SENTRY_DSN=${SENTRY_DSN:-}
+
       # Chat Configs
       - HARD_DELETE_CHATS=${HARD_DELETE_CHATS:-}
 
@@ -197,6 +200,9 @@ services:
       - LOG_DANSWER_MODEL_INTERACTIONS=${LOG_DANSWER_MODEL_INTERACTIONS:-}
       - LOG_VESPA_TIMING_INFORMATION=${LOG_VESPA_TIMING_INFORMATION:-}
 
+      # Analytics Configs
+      - SENTRY_DSN=${SENTRY_DSN:-}
+
       # Enterprise Edition stuff
       - ENABLE_PAID_ENTERPRISE_EDITION_FEATURES=${ENABLE_PAID_ENTERPRISE_EDITION_FEATURES:-false}
     extra_hosts:
@@ -254,6 +260,9 @@ services:
       - MIN_THREADS_ML_MODELS=${MIN_THREADS_ML_MODELS:-}
       # Set to debug to get more fine-grained logs
       - LOG_LEVEL=${LOG_LEVEL:-info}
+
+      # Analytics Configs
+      - SENTRY_DSN=${SENTRY_DSN:-}
     volumes:
       # Not necessary, this is just to reduce download time during startup
       - model_cache_huggingface:/root/.cache/huggingface/
@@ -283,6 +292,9 @@ services:
       # Set to debug to get more fine-grained logs
       - LOG_LEVEL=${LOG_LEVEL:-info}
       - CLIENT_EMBEDDING_TIMEOUT=${CLIENT_EMBEDDING_TIMEOUT:-}
+      
+      # Analytics Configs
+      - SENTRY_DSN=${SENTRY_DSN:-}
     volumes:
       # Not necessary, this is just to reduce download time during startup
       - indexing_huggingface_model_cache:/root/.cache/huggingface/

web/.gitignore

@@ -1,5 +1,5 @@
 # See https://help.github.com/articles/ignoring-files/ for more about ignoring files.
-
+.env.sentry-build-plugin
 # dependencies
 /node_modules
 /.pnp

web/instrumentation.ts

@@ -0,0 +1,12 @@
+export async function register() {
+  if (process.env.NEXT_PUBLIC_SENTRY_DSN) {
+    if (process.env.NEXT_RUNTIME === "nodejs") {
+      await import("./sentry.client.config");
+      await import("./sentry.server.config");
+    }
+
+    if (process.env.NEXT_RUNTIME === "edge") {
+      await import("./sentry.edge.config");
+    }
+  }
+}