From e86414a774d4a0c6db62110f04a6c96365afbf6c Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@archive.org>
Date: Sat, 30 Jan 2021 19:05:13 -0800
Subject: enable sentry exceptions for workers and pipelines

It is otherwise difficult to debug multi-million record pipelines.
---
 fatcat_scholar/work_pipeline.py | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

(limited to 'fatcat_scholar/work_pipeline.py')

diff --git a/fatcat_scholar/work_pipeline.py b/fatcat_scholar/work_pipeline.py
index aef2064..cb96274 100644
--- a/fatcat_scholar/work_pipeline.py
+++ b/fatcat_scholar/work_pipeline.py
@@ -7,11 +7,12 @@ import urllib3.exceptions
 
 import minio
 import requests
+import sentry_sdk
 import internetarchive
 from fatcat_openapi_client import ReleaseEntity, FileEntity, WebcaptureEntity
 
 from fatcat_scholar.api_entities import *
-from fatcat_scholar.config import settings
+from fatcat_scholar.config import settings, GIT_REVISION
 from fatcat_scholar.djvu import djvu_extract_leaf_texts
 from fatcat_scholar.sandcrawler import (
     SandcrawlerPostgrestClient,
@@ -469,6 +470,14 @@ def main() -> None:
         parser.print_help(file=sys.stderr)
         sys.exit(-1)
 
+    if settings.SENTRY_DSN:
+        sentry_sdk.init(
+            dsn=settings.SENTRY_DSN,
+            environment=settings.SCHOLAR_ENV,
+            max_breadcrumbs=10,
+            release=GIT_REVISION,
+        )
+
     wp = WorkPipeline(
         issue_db=IssueDB(args.issue_db_file),
         sandcrawler_db_client=SandcrawlerPostgrestClient(
-- 
cgit v1.2.3