add a last-ditch effort to find the resource in question by
listing all the resources and doing a simply match for name and
id. if no match is found then raise an error, if the list call
is unsuccessful, raise the same error. we have failed this city.
Closes-Bug: #1501362
Change-Id: I0d3d7002e9ac47b17b1ef1a5534406c85b1fc753
A few things here: 1) we need to check if the client class even
has a 'resource_class', in the case of glanceclient, it does not.
2) If everything fails we should print a better error message,
rather than a "find" failed, since some clients don't support find.
Change-Id: I6277322639e75b1635f9f3d159753efadbce1031
Instead of duplicating the same log statement throughout
the code, the same logic can be provided by a shared decorator
that abstracts away the logging capability and unifies it behind
a common function instead.
Change-Id: Icc63bced7347c8bbf0299a4c5821425a10892a79
Configuration of logging gets triggered twice. The first time it
uses the CLI options when the application is started and second
it uses the configuration file after that is read. The state of
the logging needs to be saved from the first to the second time,
so I created a class.
Implements: blueprint logging-migration
Change-Id: I7b8d1a3b6fd128e98cafd7c16009c7b694a52146
Running limits show without --absolute or --rate returns nothing and
the user is left to figure out what they need to provide to get the
correct data back. This patch prints an error and help output by making
at least one of the arguments required.
Change-Id: I576cf8ec0e05524ee67d46c48b56da8d44258667
There is no way to change the configuration variables we want
printed in log messages, so format them in the constructor.
This will save us from overridding the format method and
a couple cpu cycles every log message. This change also moves
the _LOG* variables into the formatter since they aren't really
globally needed.
Change-Id: I706e9db7da3daec20332f9d1533fe665f2739dea
Implements: blueprint logging-migration
Extract log_level from configuration file if the level was not
overridden by the command line option. The default command line
option is 1 and there is no command line option to set the
verbose_level to 1, so if it is 1, it has not be set.
Change-Id: I1be04367c72f83c1181f92ca4c2c83165b66995c
Implements: blueprint logging-migration
Move the conversion of command line options to log level out
of shell.py.
Change-Id: I86cb45a85cd63927aa1c87c1eed27542981df659
Implements: blueprint logging-migration
Add some tests for volume quota set and get rid of TODO about
using the value instead of the key to get the attribute.
Change-Id: I57aa57951aeea65965966e63af922cda532d759d
This will allow users to record logs of all their commands into
a predefined log file, in clouds.yaml. The log should have a
format similar to that of oslo.log.
Change-Id: I1b334bf429d575fc25809c9706fc0b11116be3f1
Implements: blueprint every-time-record-log-in-file
When using the command: openstack quota set, the compute quota below
can't be set successfully,the value of compute quota stay unchanged,
'fixed-ips', 'floating-ips', 'injected-files', 'key-pairs'.
What's more,I add a TODO comment in the code for two reason.
1. volume quota set works fine for the moment.
2. To indicate that this issue about volume needs
discuss and report another bug, if it's confirmed.
This bug is only about compute quota.
Change-Id: Ic1028d561f5a0030cf65ac18fc117bf01e945478
Partial-Bug: #1420104
Create a `configuration show` command that displays the current
configuration of the CLI. Different configurations can be
displayed using options such as --os-cloud. Passwords and
tokens are redacted by default unless the --unmask
option is specified.
Closes-Bug: #1476729
Change-Id: I0792365d0c5fa526cd09c0ed88c6bb1e2cb813a7
External plugins need to know which plugin interface is being
used by the host OSC. Releases <1.6 (?) will not have a
version defined.
Plugins can add the following to their make_client() to discover
the interface version:
_plugin_interface_version = getattr(
instance,
"PLUGIN_INTERFACE_VERSION",
None,
)
Change-Id: Ifc0f40fec5bc27f6425139984936b7f6e032a580
User should be able to specify the endpoint type through
a CLI optional argument/ENV variable setting. We will name this new
optional argument: --os-endpoint-type (Env: OS_ENDPOINT_TYPE) and
based on the value given, the service API will use that specific
endpoint type. Possible values: public, admin, internal.
DocImpact
Closes-Bug: #1454392
Change-Id: Ife3d4e46b44c0ddcd712b1130e27e362545a9a29
This is already fine for user_domain_id, and needs to be replicated
for project_domain_id. Also added more logging.
Change-Id: I3fa8f29edb3fc430d453bd0fc835312c0c8401f4
Adds the possibility to retrieve a project and list its
parents and subtree in the hierarchy.
Co-Authored-By: Rodrigo Duarte <rodrigods@lsd.ufcg.edu.br>
Co-Authored-By: Samuel de Medeiros Queiroz <samuel@lsd.ufcg.edu.br>
Implements: bp hierarchical-multitenancy
Change-Id: I874f6faffc8a2db9d99f12cbe0a69c0a30c0d9df
Because of the way OSC registers all plugins together we end up
with os-tenant-X parameters being saved to the project-X attribute after
parsing. If you are using the v2 plugins directly then they and os-client-config
expect the tenant_X values and will assuming no scoping information if
they are not present.
Validating options for scope will also fail in this situation, not just
because the resultant auth dictionary is missing the tenant-X
attributes, but because OSC validates that either project or domain
scope information is present.
Fix this by just always setting the v2 parameters if the v3 parameters
are present. This will have no effect on the generic or v3 case but fix
the v2 case.
Expand validation to include the tenant options so it knows that v2
plugins are scoped.
Change-Id: I8cab3e423663f801cbf2d83106c671bddc58d7e6
Closes-Bug: #1460369
This allows the server delete command to wait for the server to be
deleted (obviously).
The wait method is the same model that Tempest uses, i.e. wait for a 404
on server GET (successful deletion), fail if the server went to ERROR
status, or fail if a timeout is reached. The default timeout of 300
seconds is also what Tempest uses.
Closes-Bug: #1460112
Change-Id: I0e66c400903e82832944d1cad61e7eb30177c3e8
When we execute an Openstack CLI command, keystone should log in
Keystone access log that the user-agent that made the request was
'python-openstackclient' instead of the default 'python-keystoneclient'.
Therefore, when we create the authentication session we
need to send the explicit user-agent.
Closes-Bug: #1453995
Change-Id: I75087fd4bb1ff1e6f2a911bc70bf8008268276bb
The way that getattr is called with the None default you will always
create a compute_kwargs dictionary with key: None values. This means
that we will always send these empty requests to the servers.
Change so that only actually changed values end up in the quota set
requests and get sent.
Change-Id: I33bc3f4e1a8013ec672e995648d27513064baf26
Closes-Bug: #1451640
This adds a new option --os-cloud that allows the configuration values
for multiple clouds to be stored in a local file and selected with
a single option.
Internal option names have had 'os_' removed to be comptible with
the options returned from OpenStackConfig().get_one_cloud().
The config file is ~/.config/openstack/clouds.yaml:
Sample
------
clouds:
devstack:
auth:
auth_url: http://192.168.122.10:35357/
project_name: demo
username: demo
password: 0penstack
region_name: RegionOne
devstack:
auth:
auth_url: http://192.168.122.10:35357/
project_name: demo
username: demo
password: 0penstack
region_name: RegionOne
Co-Authored-By: Monty Taylor <mordred@inaugust.com>
Change-Id: I4939acf8067e44ffe06a2e26fc28f1adf8985b7d
Depends-On: I45e2550af58aee616ca168d20a557077beeab007
Add a --volume-type option to quota set, this will allow users
to set quotas for volume attributes on a per volume-type basis.
for example:
openstack quota set admin --volume-type myvol --volumes 12
Change-Id: I3ce9cf82a65d4f012b339f0e0dedb752cb132c33
Closes-Bug: 1438377
Subclass keystoneclient.session.Session to add the timing hooks to
record the elapsed time returned by requests.Response objects, including
the redirection history. Redirects are included individually and not
rolled into the total time for the original request.
This works for all clients that use OSC's session.
Closes-Bug: #1402577
Change-Id: I9360c90c151579b89a37edb8c11c17feb15b3cb9
Looks like the option to show limits of a specific project was
missing. This resulted in always using the authenticated project.
Change-Id: I512a05df20860ffb52af34f3b64c9eb81dae8c61
Closes-Bug: 1438379
The plugin detection at the top of openstackclient.api.auth did not
detect the plugins at the bottom of that file because, surprise,
they had not been declared yet so the entry points were ignored.
Move both plugin subclasses into openstackclient.api.auth_plugin.
Fix a problem with the password callback that was otherwise made
worse with this change.
Closes-Bug: 1428912
Change-Id: Idc3b72534071e0013c8922884a8bc14137509a0f
Not returning a value is the same as returning None. In the event that
someone asks ClientManager for an attribute that doesn't exist it should
raise AttributeError in the same way as other python objects rather
than return an empty value.
Change-Id: Id0ee825e6527c831c38e3a671958ded362fb96e1
