Merge pull request #1937 from dandi/sha256-checksum-cron-job

Add retries to sha256 checksum calculation task

Author: jjnesbitt

dandiapi/api/tasks/__init__.py

@@ -1,6 +1,9 @@
 from __future__ import annotations
 
+from typing import TYPE_CHECKING
+
 from celery import shared_task
+from celery.exceptions import SoftTimeLimitExceeded
 from celery.utils.log import get_task_logger
 
 from dandiapi.api.doi import delete_doi
@@ -15,6 +18,9 @@
 from dandiapi.api.models import Asset, AssetBlob, Version
 from dandiapi.api.models.dandiset import Dandiset
 
+if TYPE_CHECKING:
+    from uuid import UUID
+
 logger = get_task_logger(__name__)
 
 
@@ -26,10 +32,16 @@ def remove_asset_blob_embargoed_tag_task(blob_id: str) -> None:
     remove_asset_blob_embargoed_tag(asset_blob)
 
 
-@shared_task(queue='calculate_sha256', soft_time_limit=86_400)
-def calculate_sha256(blob_id: str) -> None:
+@shared_task(
+    queue='calculate_sha256',
+    soft_time_limit=86_400,  # 24 hours
+    autoretry_for=(SoftTimeLimitExceeded,),
+    retry_backoff=True,
+    max_retries=3,
+)
+def calculate_sha256(blob_id: str | UUID) -> None:
     asset_blob = AssetBlob.objects.get(blob_id=blob_id)
-    logger.info('Found AssetBlob %s', blob_id)
+    logger.info('Calculating sha256 checksum for asset blob %s', blob_id)
     sha256 = asset_blob.blob.storage.sha256_checksum(asset_blob.blob.name)
 
     # TODO: Run dandi-cli validation