1. upload a large image, for example 50G
2. kill glance-api when image status:saving
3. restart glance-api
4. delete image
the image content that have been uploaded can not be deleted. I think the add method should put the object manifest onto swift first, before we upload the content when we upload a large image in chunks.
manifest = "%s/%s-" % (location.container, location.obj)
headers = {'ETag': hashlib.md5("").hexdigest(), 'X-Object-Manifest': manifest}
connection.put_object(location.container, location.obj, None, headers=headers)
the code above shoud put before the code we upload the image chunks.
1. upload a large image, for example 50G
2. kill glance-api when image status:saving
3. restart glance-api
4. delete image
the image content that have been uploaded can not be deleted. I think the add method should put the object manifest onto swift first, before we upload the content when we upload a large image in chunks. container, location.obj) md5("") .hexdigest( ), 'X-Object- Manifest' : manifest} put_object( location. container, location.obj, None, headers=headers)
manifest = "%s/%s-" % (location.
headers = {'ETag': hashlib.
connection.
the code above shoud put before the code we upload the image chunks.