vbatts/quay
Archived
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

Change manifest backfill to lookup missing content checksums in the repository

Browse source 
After discussion, we decided the best solution for the missing content checksum problem was to lookup the proper blobs in the repository and, if not present, mark the manifest as broken, as this would reflect the actual issue the user faces if they pull the repository tag today via V2
This commit is contained in:
Joseph Schorr 2018-08-14 14:36:19 -04:00
parent c0d25d3149
commit 70df10dade
3 changed files with 92 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

23
workers/manifestbackfillworker.py
View file

@ -8,6 +8,7 @@ from data.database import (UseThenDisconnect, TagManifest, TagManifestToManifest
db_transaction)
from data.model.image import get_parent_images
from data.model.tag import populate_manifest
from data.model.blob import get_repo_blob_by_digest, BlobDoesNotExist
from image.docker.schema1 import (DockerSchema1Manifest, ManifestException, ManifestInterface,
DOCKER_SCHEMA1_SIGNED_MANIFEST_CONTENT_TYPE)
from workers.worker import Worker
@ -127,10 +128,28 @@ def backfill_manifest(tag_manifest):
root_image = tag_manifest.tag.image
repository = tag_manifest.tag.repository
storage_ids = {root_image.storage.id}
image_storage_id_map = {root_image.storage.content_checksum: root_image.storage.id}
parent_images = get_parent_images(repository.namespace_user.username, repository.name, root_image)
for parent_image in parent_images:
storage_ids.add(parent_image.storage.id)
image_storage_id_map[parent_image.storage.content_checksum] = parent_image.storage.id
# Ensure that all the expected blobs have been found. If not, we lookup the blob under the repo
# and add its storage ID. If the blob is not found, we mark the manifest as broken.
storage_ids = set()
for blob_digest in manifest.blob_digests:
if blob_digest in image_storage_id_map:
storage_ids.add(image_storage_id_map[blob_digest])
else:
logger.debug('Blob `%s` not found in images for manifest `%s`; checking repo',
blob_digest, tag_manifest.id)
try:
blob_storage = get_repo_blob_by_digest(repository.namespace_user.username, repository.name,
blob_digest)
storage_ids.add(blob_storage.id)
except BlobDoesNotExist:
logger.debug('Blob `%s` not found in repo for manifest `%s`',
blob_digest, tag_manifest.id)
is_broken = True
with db_transaction():
# Re-retrieve the tag manifest to ensure it still exists and we're pointing at the correct tag.