| *** diablo_rojo is now known as Guest6733 | 01:00 | |
| *** sshnaidm is now known as sshnaidm|off | 08:04 | |
| gtema | thomasb06 - if your question is still valid - you should not really use git pull (unless willing to go to real clean master). Instead you do `git review -d CHANGE_ID`. This will ensure you get latest version of your change | 13:21 |
|---|---|---|
| gtema | otherwise (for branch) `git checkout master && git pull` | 13:21 |
| opendevreview | Alfredo Moralejo proposed openstack/python-openstackclient master: Replace assertItemsEqual with assertCountEqual https://review.opendev.org/c/openstack/python-openstackclient/+/808079 | 13:50 |
| ebbex | I have a problem where "openstack image set --property hw_qemu_guest_agent=no <uuid> && openstack image show <uuid>" always shows up as hw_qemu_guest_agent='True' in the output, can anyone else see if they get the same result? | 14:52 |
| gtema | you may try to use `openstack image unset --property hw_qemu_guest_agent <uuid>`. Maybe this can make it | 14:53 |
| ebbex | "glance image-show <uuid>" shows it correctly as "no". | 14:53 |
| ebbex | I can even set the property to "horse" and glance shows it correctly but the openstack-client always evaluates it to 'True' for some reason. | 14:54 |
| gtema | hmm, can you post output of the API response from`openstack image show <uuid> --debug`? | 14:54 |
| ebbex | and it's only this property for some reason, hw_machine_type, hw_disk_bus, hw_scsi_model, etc. all fine. | 14:55 |
| gtema | hmm | 14:55 |
| gtema | ok, found | 14:56 |
| gtema | in the SDK the attribute is declared as bool | 14:56 |
| gtema | which means it will not really support `no`, but `false` | 14:56 |
| ebbex | aha, that explains that I suppose. | 14:59 |
| ebbex | the docs I've read on the property say it should be "yes", so really scratched my head when I set it to "no" but it still showed as 'True'. | 15:00 |
| gtema | that was quite bad implementation decision in glance/wherever, cause all other known attributes are really bools, here it is described as "boolean `yes` or `no`" | 15:03 |
| ebbex | hmm, well it doesn't help that i set the property to 'False', 'false' or 'no'. It still reports it back as 'True'. | 15:09 |
| gtema | well: | 15:11 |
| * gtema < https://matrix.org/_matrix/media/r0/download/matrix.org/HdLyfeqfzrrPMQcOLSPvdAMY/message.txt > | 15:11 | |
| ebbex | oh dear :( | 15:13 |
| ebbex | so there is only setting the property or unsetting it? But nova respects the property if 'no' or 'yes'? (I'll have to check that out next) | 15:14 |
| gtema | this question I can not answer sadly | 15:15 |
| ebbex | nova/objects/image_meta.py: 'hw_qemu_guest_agent': fields.FlexibleBooleanField() | 15:16 |
| ebbex | so nova seems to respect the the value, they even have tests for it with values "yes" and "no". | 15:19 |
| gtema | https://opendev.org/openstack/oslo.utils/src/branch/master/oslo_utils/strutils.py#L122 | 15:20 |
| gtema | this is where it lands | 15:20 |
| gtema | anyway - we should fix it in SDK. This is clearly not a bool, but string for us | 15:20 |
| ebbex | thanks, any place I should file a bug for this? | 15:22 |
| gtema | reports are not welcome, but fixes are ;-) | 15:23 |
| gtema | I will do this shortly | 15:23 |
| opendevreview | Merged openstack/openstacksdk master: Add compute microversion 2.78 https://review.opendev.org/c/openstack/openstacksdk/+/786290 | 15:24 |
| ebbex | Is it as easy as changing the line "hw_qemu_guest_agent = resource.Body('hw_qemu_guest_agent', type=bool)" ? | 15:24 |
| ebbex | (to something else that is) | 15:25 |
| opendevreview | Artem Goncharov proposed openstack/openstacksdk master: Change image.hw_qemu_guest_agent to be string https://review.opendev.org/c/openstack/openstacksdk/+/808101 | 15:31 |
| gtema | yeah ebbex - like that but also test should be adapted | 15:31 |
| *** efried1 is now known as efried | 17:30 | |
| *** slaweq1 is now known as slaweq | 19:19 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!