Faster cache lookup by removing a join with the ImagePlacementTable, removing the extra loop to add the locations and filtering the images looked up by the base image

This commit is contained in:
Joseph Schorr 2015-04-24 16:22:19 -04:00
parent fd65ca5916
commit e70343d849
3 changed files with 39 additions and 1 deletions

View file

@ -104,7 +104,9 @@ class BuildJob(object):
return None return None
# Build an in-memory tree of the full heirarchy of images in the repository. # Build an in-memory tree of the full heirarchy of images in the repository.
all_images = model.get_repository_images(repo_namespace, repo_name) all_images = model.get_repository_images_directly(repo_build.repository,
with_ancestor=base_image)
all_tags = model.list_repository_tags(repo_namespace, repo_name) all_tags = model.list_repository_tags(repo_namespace, repo_name)
tree = ImageTree(all_images, all_tags, base_filter=base_image.id) tree = ImageTree(all_images, all_tags, base_filter=base_image.id)

View file

@ -1751,6 +1751,21 @@ def get_matching_repository_images(namespace_name, repository_name, docker_image
return _get_repository_images_base(namespace_name, repository_name, modify_query) return _get_repository_images_base(namespace_name, repository_name, modify_query)
def get_repository_images_directly(repository, with_ancestor=None):
query = (Image
.select(Image, ImageStorage)
.join(ImageStorage)
.where(Image.repository == repository))
if with_ancestor:
ancestors_string = '%s%s/' % (with_ancestor.ancestors, with_ancestor.id)
query = query.where((Image.ancestors ** (ancestors_string + '%')) |
(Image.id == with_ancestor.id))
return query
def get_repository_images(namespace_name, repository_name): def get_repository_images(namespace_name, repository_name):
return _get_repository_images_base(namespace_name, repository_name, lambda q: q) return _get_repository_images_base(namespace_name, repository_name, lambda q: q)

View file

@ -91,6 +91,27 @@ class TestImageTree(unittest.TestCase):
self.assertEquals('staging', tree.tag_containing_image(result[0])) self.assertEquals('staging', tree.tag_containing_image(result[0]))
def test_longest_path_simple_repo_direct_lookup(self):
repository = model.get_repository(NAMESPACE, SIMPLE_REPO)
all_images = list(model.get_repository_images(NAMESPACE, SIMPLE_REPO))
all_tags = list(model.list_repository_tags(NAMESPACE, SIMPLE_REPO))
base_image = self._get_base_image(all_images)
tag_image = all_tags[0].image
def checker(index, image):
return True
filtered_images = model.get_repository_images_directly(repository, with_ancestor=base_image)
self.assertEquals(set([f.id for f in filtered_images]), set([a.id for a in all_images]))
tree = ImageTree(filtered_images, all_tags)
ancestors = tag_image.ancestors.split('/')[2:-1] # Skip the first image.
result = tree.find_longest_path(base_image.id, checker)
self.assertEquals(3, len(result))
self.assertEquals('latest', tree.tag_containing_image(result[-1]))
if __name__ == '__main__': if __name__ == '__main__':
unittest.main() unittest.main()