| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2020-06-18 | |||
| 12:13:12 | mordred | slaweq: see comment from frickler on https://review.opendev.org/#/c/736005/ - it's possible there is a neutron bug | |
| 12:13:30 | slaweq | mordred: looking | |
| 12:14:27 | mordred | slaweq: (it's possible it's not - but when I see "unintended information leak" ...) | |
| 12:34:09 | slaweq | mordred: I just opened https://bugs.launchpad.net/neutron/+bug/1884067 | |
| 12:34:09 | openstack | Launchpad bug 1884067 in neutron "[API] Filtering by fields not allowed to see is possible for regular users" [High,Confirmed] | |
| 12:34:16 | slaweq | mordred: frickler thx for finding this | |
| 12:35:19 | slaweq | mordred: frickler I think we are not checking what fields are available for user during filtering, we are only removing such fields from the returned data | |
| 12:35:23 | frickler | slaweq: sorry, I wanted to ping you earlier about this but then got distracted, it just looked odd to me when testing mnaser's patch | |
| 12:35:35 | slaweq | I will ask amotoki to take a look into it | |
| 12:38:00 | mordred | \o/. we're helpful! good catch frickler :) | |
| 12:38:23 | mordred | mnaser: you made things betterer | |
| 12:39:28 | mnaser | what did I do oops | |
| 12:39:55 | mnaser | oh no | |
| 12:39:58 | mnaser | that’s no good | |
| 12:40:19 | mnaser | we should fix that soon :) | |
| 12:40:48 | mnaser | (and if no one has the time to do that, I can, cause you know, that’s a little important) | |
| 12:41:14 | mnaser | I’m not that worried though, we don’t expose the host id anywhere so I don’t think anyone can see it to use it later | |
| 13:51:31 | shaohe_feng | hi, mordred and gtema | |
| 13:52:01 | shaohe_feng | there is a issue when I connect glance from cyborg | |
| 13:52:11 | shaohe_feng | Need you help | |
| 13:53:30 | shaohe_feng | Traceback (most recent call last): | |
| 13:53:31 | shaohe_feng | "/opt/stack/cyborg/cyborg/image/api.py", line 163, in download | |
| 13:56:10 | mordred | I want to help shaohe_feng | |
| 13:56:26 | shaohe_feng | Traceback (most recent call last): | |
| 13:56:27 | shaohe_feng | "/opt/stack/cyborg/cyborg/image/api.py", line 163, in download | |
| 14:12:29 | mordred | shaohe_feng: try pasting to paste.openstack.org and then copying the link here | |
| 14:14:02 | shaohe_feng | mordred: this is the traceback http://paste.openstack.org/show/794946/ | |
| 14:17:00 | shaohe_feng | mordred: this is the conf http://paste.openstack.org/show/794948/ | |
| 14:19:48 | shaohe_feng | mordred: this is the cyborg repo https://github.com/openstack/cyborg and glance api code is here: https://github.com/openstack/cyborg/blob/6dab512af226677c9065f85f264a691f3aeb6eab/cyborg/image/glance.py | |
| 14:20:25 | mordred | shaohe_feng: sadly I don't think we're going to be much help with that - it's using glanceclient not openstacksdk - I'm really not sure what's going on :( | |
| 14:21:05 | shaohe_feng | mordred I'm using SDK: | |
| 14:21:48 | shaohe_feng | + ksa_adap = utils.get_sdk_adapter( | |
| 14:22:12 | shaohe_feng | in my code I chage it to get_sdk_adapter | |
| 14:23:26 | shaohe_feng | here I change get_ksa_adapter to get_sdk_adapter https://github.com/openstack/cyborg/blob/6dab512af226677c9065f85f264a691f3aeb6eab/cyborg/image/glance.py#L113 | |
| 14:24:33 | mordred | shaohe_feng: yes - but the traceback you pasted is flowing through glanceclient - so it seems that sdk adapter is just being used to find the list of glance servers | |
| 14:25:09 | mordred | so - at some point that info is being passed to a Glanceclient constructor - and god only knows what it's doing | |
| 14:25:33 | shaohe_feng | https://github.com/openstack/cyborg/blob/master/cyborg/common/utils.py#L135 | |
| 14:25:44 | shaohe_feng | ^ mordred: what the get_sdk_adapter define | |
| 14:26:00 | shaohe_feng | so why it still use glanceclient ? | |
| 14:26:14 | shaohe_feng | I already change it to get_sdk_adapter | |
| 14:26:54 | shaohe_feng | something wrong when I change it to get_sdk_adapter ？ | |
| 14:27:27 | shaohe_feng | mordred can you tell me what should I do? | |
| 14:28:02 | mordred | shaohe_feng: that looks fine - but there is still this: https://github.com/openstack/cyborg/blob/6dab512af226677c9065f85f264a691f3aeb6eab/cyborg/image/glance.py#L129 | |
| 14:28:11 | mordred | shaohe_feng: let me look at the code a little more and see if I can make some suggestions | |
| 14:28:38 | shaohe_feng | mordred: thank you. | |
| 14:30:07 | gtema | shaone_feng - I guess you should be targeting dropping your cyborg/image/glance completely and modify cyborg/image/api to use SDK instead | |
| 14:31:19 | mordred | gtema: looking at the master branch, it looks like they're laready doing that | |
| 14:31:43 | gtema | oh, I was looking to the posted particular commit | |
| 14:31:45 | mordred | oh - wait | |
| 14:31:47 | mordred | eah | |
| 14:32:22 | mordred | ZOMG - the method is returning a GlanceClient and calling it a session | |
| 14:32:25 | gtema | even in master there is still from cyborg.image import glance | |
| 14:33:05 | mordred | yup. | |
| 14:33:13 | mordred | I was fooled by "session.delete()" | |
| 14:38:02 | shaohe_feng | gtema: mordred: I'm still use GlanceClient ? | |
| 14:39:45 | mordred | gtema: they're going to nerdsnipe me in to fixing cyborg aren't they? | |
| 14:42:46 | gtema | shaone_feng - you stacktrace exactly points into that | |
| 14:42:56 | gtema | mordred - right | |
| 14:42:59 | shaohe_feng | gtema: mordred: sorry, I'm offline just now. something wrong with my network. | |
| 14:43:11 | shaohe_feng | may I missing something? | |
| 14:43:40 | gtema | shaohe_feng - guess not | |
| 14:43:52 | mordred | shaohe_feng: cyborg.image.api.API._get_session_and_image_id is constructing a Glanceclient object | |
| 14:44:05 | mordred | shaohe_feng: which in turn is used in API.download | |
| 14:44:06 | shaohe_feng | good, so what should I change? | |
| 14:44:23 | mordred | probably the entire cyborg/image/glance.py file | |
| 14:44:31 | shaohe_feng | the FPGA need to download image | |
| 14:44:40 | gtema | yes - drop it and modify the cyborg/image/api | |
| 14:44:42 | shaohe_feng | we need to program the FPGA | |
| 14:44:53 | shaohe_feng | any example for it? | |
| 14:44:55 | mordred | shaohe_feng: I'm going to see if I can make a non-working example patch that can point you in the right direction | |
| 14:45:10 | shaohe_feng | mordred good. thanks. | |
| 14:45:34 | mordred | gtema: it's the same code as from nova, so if we fix it here for cyborg, we'll have a good example of how to fix it for nova too :) | |
| 14:45:49 | gtema | :) | |
| 16:02:45 | elmiko | API SIG office hour open | |
| 16:04:39 | elmiko | mordred: i haven't seen much more activity on the service discovery PRs, i'm going to move towards a freeze and asking for wider review if that sounds good to you? | |
| 16:07:31 | mordred | elmiko: ++ | |
| 16:07:46 | elmiko | thank you sir =) | |
| 16:09:45 | mordred | thank you! | |
| 16:34:08 | elmiko | ok, frozen and email sent | |
| 16:51:00 | dtantsur | \o/ | |
| 17:08:33 | elmiko | have a good weekend all, stay safe and healthy out there =) o/ | |
| 22:06:32 | openstackgerrit | Merged openstack/keystoneauth stable/ussuri: Fix E741 pep8 failure https://review.opendev.org/731188 | |
| #openstack-sdks - 2020-06-19 | |||
| 01:41:33 | openstackgerrit | Lance Bragstad proposed openstack/keystoneauth stable/ussuri: Inject /v3 in token path for v3 plugins https://review.opendev.org/731519 | |
| 07:30:21 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: Add support for filters into the project cleanup https://review.opendev.org/735336 | |
| 07:42:02 | openstackgerrit | Sagi Shnaidman proposed openstack/ansible-collections-openstack master: New CI jobs configuration https://review.opendev.org/736946 | |
| 07:46:34 | openstackgerrit | Sagi Shnaidman proposed openstack/ansible-collections-openstack master: New CI jobs configuration https://review.opendev.org/736946 | |
| 07:47:55 | openstackgerrit | Sagi Shnaidman proposed openstack/ansible-collections-openstack master: New CI jobs configuration https://review.opendev.org/736946 | |
| 07:53:25 | openstackgerrit | Sagi Shnaidman proposed openstack/ansible-collections-openstack master: New CI jobs configuration https://review.opendev.org/736946 | |
| 08:32:42 | openstackgerrit | Artem Goncharov proposed openstack/ansible-collections-openstack master: Add more useful information from exception https://review.opendev.org/736953 | |
| 10:46:17 | openstackgerrit | Jens Harbott (frickler) proposed openstack/python-openstackclient master: WIP: Allow more options when creating a router https://review.opendev.org/736973 | |
| 12:45:51 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: Add support for filters into the project cleanup https://review.opendev.org/735336 | |
| 15:56:05 | AJaeger | could I get a review for https://review.opendev.org/#/c/734852/ - updating docstheme for service-types-authority, please? | |
| 15:56:16 | AJaeger | mordred ^ | |
| 15:58:06 | mordred | AJaeger: done | |
| 15:58:55 | AJaeger | thanks! | |
| 15:59:00 | gtema_ | mordred, do you remember what is this networking problem that blocks everything? | |
| 16:09:33 | openstackgerrit | Merged openstack/service-types-authority master: Switch to newer openstackdocstheme version https://review.opendev.org/734852 | |
| 17:39:41 | holser | mordred Quick question. Why do we install python-dev on Ubuntu and don't install it for CentOS/RedHat? I am talking about https://opendev.org/zuul/zuul-jobs/src/branch/master/roles/ensure-python/tasks/main.yaml#L9 | |
| 17:48:50 | holser | mordred opened https://storyboard.openstack.org/#!/story/2007838 | |
| 18:35:19 | openstackgerrit | Sagi Shnaidman proposed openstack/ansible-collections-openstack master: New CI jobs configuration https://review.opendev.org/736946 | |