Correct content type of response for image metadata getting api

[zhiyanliu]

The recent change [0] makes glanceclient use requests to communicate
with glance server, new http handling logic of which is sensitive on the
content type of response [1].

But current jumpgate logic for the api doesn't return any content by
http body [2] but with 'application/json' content type [3]. Then it
broken that new glanceclient logic and trigger error: "Multiple image
matches found for '<IMAGE_ID>', use an ID to be more specific.".

[0]
openstack/python-glanceclient@dbb242b
[1]
https://github.com/openstack/python-glanceclient/blob/master/glanceclient/common/http.py#L218
[2]
https://github.com/softlayer/jumpgate/blob/master/jumpgate/image/drivers/sl/images.py#L223
[3] https://github.com/racker/falcon/blob/master/falcon/__init__.py#L27

Signed-off-by: Zhi Yan Liu zhiyanl@cn.ibm.com

[jimlindeman]

+1
So essentially all we're fixing here is the HEAD rest call for a specific image to just return back a content-type of plain "text/html" so it doesn't trigger the default of "application/json". I see that at http://www.packetizer.com/ws/rest.html, it says this for the HEAD call:
HEAD is similar to GET. The significant difference is that the representation is not returned.