The client should check the update request before sending it to the server

[Flavio Percoco]

[root@orange-vdse ~(keystone_admin)]# glance --os-image-api-version 2 image-update fc4bfa83-6eea-4090-91a3-92b02668570d --owner 637ea20d137e4fa3bdbbb89ec513ef05Request returned failure status.
403 Forbidden
Attribute 'owner' is reserved.
    (HTTP 403)

However, if the owner remains unchanged, the client will send the request
to the server and not raise an error.

[fedora@cinder devstack]$ glance --os-image-api-version 2 image-update fc4bfa83-6eea-4090-91a3-92b02668570d --owner cf7d037795cd45919e9d2824c72d0bae
+------------------+--------------------------------------+
| Property | Value |
+------------------+--------------------------------------+
| checksum | 51bc16b900bf0f814bb6c0c3dd8f0790 |
| container_format | bare |
| created_at | 2014-05-02T10:36:23Z |
| disk_format | qcow2 |
| id | fc4bfa83-6eea-4090-91a3-92b02668570d |
| min_disk | 0 |
| min_ram | 0 |
| name | Fedora-x86_64-20-20131211.1-sda |
| owner | cf7d037795cd45919e9d2824c72d0bae |
| protected | False |
| size | 214106112 |
| status | active |
| tags | [] |
| updated_at | 2014-05-02T10:36:24Z |
| visibility | public |
+------------------+--------------------------------------+

[Flavio Percoco (flaper87)]

turns out the client is not sending the owner field update, which IMHO, is wrong. These kind of optimizations belong to the server, not the client. I'll move this bug to the client and update the title.

[Erno Kuvaja (jokke)]

IMO the client should tell the user that the operation was not attempted due to no-change situation and fail early as we do in other similar cases (image create with known not working conditions, etc.). I don't think, we should send request to the server where we know already that the server is not going to do anything about it.

[Cindy Pallares (cpallares)]

Erno, perhaps there is something I am missing, but how should the client know it is a no-change situation? Doesn't it have to send the request to the server and get a response to compare?

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/185223

[OpenStack Infra (hudson-openstack)]

Change abandoned by Cindy Pallares (<email address hidden>) on branch: master
Review: https://review.openstack.org/99413

[Erno Kuvaja (jokke)]

Hi Cindy,

I was wrong about our current state and how we did that cleanup of those parameters. We had lengthy conversation with Flavio around this and what we probably should do is clean out that whole fetching the object beforehand and sanitizing our request from that cache and rather just format the request based on the schema and send it out to the server.

Let me know what do you think about that.

Honestly I did not read the change itself and my comments were based on the bug and the commit message. I'm sorry I missed your response.

[OpenStack Infra (hudson-openstack)]

Change abandoned by Glance Bot (<email address hidden>) on branch: master
Review: https://review.openstack.org/185223

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/302572

[OpenStack Infra (hudson-openstack)]

Change abandoned by Brian Rosmaita (<email address hidden>) on branch: master
Review: https://review.openstack.org/302572
Reason: No activity in >1 year.