Public bug reported:
if no container name is provided, horizon ends up passing "" to the
cinder client api call. This is an invalid container name - None is the
correct value to pass for this to work as desired.
** Affects: cinder
Importance: Undecided
Status: New
** Affects:
Public bug reported:
When trying to delete a container within swift, we get a pop-up to
confirm we want it deleted.
If you act very quickly, you can confirm "yes" and delete the container.
If you are slow, you get redirected back to the overview / home page.
** Affects: horizon
Importance:
Public bug reported:
>From time to time, the left hand nav bar does not display correctly.
This seems to occur a bit more frequently when you switch projects with
different roles (admin to _member_ or back and fort), and also occurs
when horizon is provided by several backend servers (behind a
Public bug reported:
The login page "Sign In" button is not showing up. It looks like this
was introduced with this fix:
https://review.openstack.org/#/c/376064/
Proposing to revert that patch.
** Affects: horizon
Importance: Critical
Assignee: Eric Peterson (eric
Public bug reported:
The angular images panel does not always show deactivated image status
correctly.
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: In Progress
--
You received this bug notification because you are a member
Public bug reported:
The button icons use + / - and really the arrow buttons are more
appropriate for moving items up/down.
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: In Progress
** Changed in: horizon
Assignee: (unassigned
Public bug reported:
Trying to list private images on the glance / images panel yields an
error.
It looks like Private should actually be private instead.
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: New
** Changed in: horizon
Public bug reported:
running "tox -e cover" is now yielding this failure:
OK (SKIP=8)
Destroying test database for alias 'default'...
cover runtests: commands[3] | coverage combine
cover runtests: commands[4] | coverage xml
No data to report.
ERROR: InvocationError:
Public bug reported:
The default policy for this call:
novaclient(request).quotas.get(tenant_id, detail=True)
fails unless I am an admin type user. This doesn't seem to make sense,
as an _member_ type user, I can still find all the details just the
same. This just makes user do many more
Public bug reported:
if you use the material theme and then try to add a _footer.html file
for a consistent footer, the content is not laid out well at all.
There are docs on how to add a simple _footer.html, and this should not
require heavy lifting in sccs etc just to get some footer content
Public bug reported:
https://bugs.launchpad.net/horizon/+bug/1398845 Broke network creation
for normal users, that also want to create a subnet.
The policy check is not the correct one to use, as there is no existing
network to check ownership against.
** Affects: horizon
Importance:
Public bug reported:
The base form class has a form_invalid implementation. This class calls
the super get context data, which then makes it so my form does not have
the context data I would normally expect to have.
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson
Public bug reported:
Seeing that the new swift ui cannot handle duplicate pseudo folder names
:
d8f1634c7721.js:1235 Error: [ngRepeat:dupes] Duplicates in a repeater
are not allowed. Use 'track by' expression to specify unique keys.
Repeater: item in items track by item[config.trackId],
Public bug reported:
When a user has the admin role, the count of used subnets is that of all
subnets in existence (in that region).
This makes it so admin users cannot create subnets.
The other parts of the quota code does some filtering to ensure only
resources for the current project are
Public bug reported:
When creating a security group rule, we have help text that says -1 is a
valid option.
We have a validator that does not allow -1.
The neutron client also does not support -1.
I believe we just need help text corrected.
** Affects: horizon
Importance: Undecided
as
ideal.
Horizon should add the option of having the nova scheduler place the VM.
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: In Progress
--
You received this bug notification because you are a member of Yahoo!
Engineering Team, which
Public bug reported:
We are unable to make any changes for a project, this is related to the
quotas and calculation of quotas for subnets.
As an admin, the subnet list count is giving me a high number, say 1000.
This is because it is listing all subnets everywhere.
When I try to make changes to
Public bug reported:
We have users remove keypairs a little too freely. We would like to
improve the error message / warning text for users.
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: In Progress
** Changed in: horizon
Public bug reported:
This bug is not 100% reliable to reproduce, and occurs only when running
under wsgi / apache in my testing. Developers will not typically see
this bug.
When wsgi is tuned to have multiple processes, the heat template upload
can have parse failures. These errors do not show
Public bug reported:
My users have been complaining that horizon's ui for managing volume
attachment to vms is very strange / tough to use. Would like to see
some other approach / ideas.
** Affects: horizon
Importance: Undecided
Status: New
--
You received this bug notification
Public bug reported:
Recent horizon changes
https://github.com/openstack/horizon/commit/f38ddf20323ca1e278caa37013871b65013a0e81#diff-15e3003726d7b90816528ee8248b000f
as well as some likely changes with the heat client have made it so the heat
resource drill down page does some really awful
Public bug reported:
The detailed response from the nova usage call, reports invalid instance
size info if the instance has been resized. This API is used by
Horizon, which is how we found this.
Steps to reproduce:
1) Create an instance, say small
2) resize the instance, something like large
Public bug reported:
When creating a subnet, users currently have a blank screen presented to
them for dns servers.
For our deployment, we would like to have a configuration setting where
we can provide some default dns servers to use.
This seems like a common need, for operators to ask that
Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: In Progress
** Changed in: horizon
Assignee: (unassigned) => Eric Peterson (ericpeterson-l)
--
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenSt
stuff.
** Affects: horizon
Importance: Medium
Assignee: Eric Peterson (ericpeterson-l)
Status: Confirmed
** Changed in: horizon
Assignee: (unassigned) => Eric Peterson (ericpeterson-l)
--
You received this bug notification because you are a member of Yahoo!
Engineer
Public bug reported:
We have instances which have multiple ports, one port goes into the
tenant network - and the other port goes into a network which is not
visible to the tenant. ( This other network is used as a control plane
for lbaas management)
With these instances, they only *sometimes*
Public bug reported:
The check marks for selected project etc do not display well in safari /
mac. They essentially get placed on top of the project name, and it
looks bad.
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: In Progress
Public bug reported:
If a deployment has defined volume types, having "No volume type" show
up as the first item in the list is not a great experience.
Instead:
"No volume type" should show up when cinder has 0 volume types defined.
-otherwise-
Use the list from cinder's api call.
** Affects:
Public bug reported:
There was a small change to add a required field for launch instance:
https://bugs.launchpad.net/horizon/+bug/1283148
This new config setting should be used within ng launch instance as
well.
** Affects: horizon
Importance: Medium
Status: Confirmed
--
You
Public bug reported:
We have a setup with 3 control nodes, and each tenant network gets 2
dhcp agents. Most of the time, this works fine / well.
However, we have seen where the dhcp agents assigned to a network might
point to control nodes 1 2 - but the port listing for the tenant
network has
** Changed in: horizon
Status: Confirmed = Fix Released
--
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenStack Dashboard (Horizon).
https://bugs.launchpad.net/bugs/1212358
Title:
django openstack auth is granting
Public bug reported:
When the data table displayed requires vertical scrolling, the left hand
nav's background coloring cuts off eventually as it is sized to 100%
of the display area.
It looks like:
http://pasteboard.co/2Ni16M2C.png
** Affects: horizon
Importance: Undecided
Public bug reported:
The tactic of using the overrides file vs the enabled folder vs other
means is unclear to plugin developers.
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: In Progress
** Changed in: horizon
Assignee
Public bug reported:
We have disabled old launch instance, and enabled the new launch
instance.
When we access the network topo page, we see two launch instance
buttons.
** Affects: horizon
Importance: Undecided
Status: New
** Tags: low-hanging-fruit
** Tags added:
Public bug reported:
The lbaas table actions all use a post() call with a lot of if/then/else
blocks checking the actions. This makes extending this page very
tricky, and does not follow documentation and examples / best practices
of how to implement table actions in a more flexible / extensible
Public bug reported:
I have 3 identical machines (host names are different), and I have an
ansible script to deploy horizon. As part of the script we run
collectstatic and compress by hand (offline compression).
Each machine's static contents looks slightly different.
This causes horizon to
Public bug reported:
there are lots of changes recently, especially around
fd.populate_horizon_config() calls
deployers find the need to make changes to settings.py and
local_settings.py, because of what can be done where.
The churn in settings.py should be minimized, especially around
Public bug reported:
When the static url path is something like:
/static
And the dashboard path is something like
/horizon
or
/dashboard
Then the old launch instance works fine and other js libraries are also
ok/fine.
When I configure horizon to use the new launch instance screen (and
disable
Public bug reported:
The launch instance flow has a config control for
OPENSTACK_HYPERVISOR_FEATURES['can_set_password'] which allows a
deployment to decide if the password should be prompted for during
instance creation.
If a password is not possible to set, then it is very likely that an ssh
Public bug reported:
the local_settings.py example file redefines the HORIZON_CONFIG section
of config info.
This creates confusion as to what is used why and where to make changes.
The correct way is to have the local_settings.py example do things like:
HORIZON_CONFIG[password_validator]
or
Public bug reported:
See this screen shot:
http://pasteboard.co/14TpE09S.png
The longer role names or having multiple roles is not working well
beyond the most basic _member_ admin type roles.
** Affects: horizon
Importance: Undecided
Status: New
** Tags: low-hanging-fruit ops
Public bug reported:
Seeing that the table action allowed() call (for a table action, not a
row action) has an exception / error - and then makes a call to the
horizon messages.warning() api. The actual error is not showing until
the subsequent page load call. This makes it look like the
Public bug reported:
I am seeing the images table fill up our log file with messages that are
completely useless / silly. There should be a getattr() with a default
to None so this type of situation is more gracefully handled.
2015-06-01 20:50:11,741 32061 WARNING horizon.tables.base
Public bug reported:
The floating IP associated with a load balancers vip is perhaps one of
the most important pieces of information users need to know.
Horizon has no clear / easy way to get this information other than
tracking down a VIP's attached port, and mapping that to the FIP to port
Public bug reported:
Seeing this problem with recent DOA, against keystone v2:
File
/foo/current/.venv/lib/python2.7/site-packages/openstack_auth/utils.py, line
54, in middleware_get_user
request._cached_user = get_user(request)
File
Public bug reported:
The pool creation UI has a selection for choosing the subnet. What is
shown is the IP range, something like:
192.168.1.0/24 (whatever)
The problem with this is that I could use multiple subnets with the same
or very similar IP ranges. The UI is not providing any clue as
Public bug reported:
The lbaas add vip screen has a dropdown of subnets to select. Right now
only the cidr is displayed, which is not always helpful.
I'd like to see the subnet name displayed as well.
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l
: Eric Peterson (ericpeterson-l)
Status: New
** Changed in: horizon
Assignee: (unassigned) = Eric Peterson (ericpeterson-l)
--
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenStack Dashboard (Horizon).
https
: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: New
** Changed in: horizon
Assignee: (unassigned) = Eric Peterson (ericpeterson-l)
--
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed
Public bug reported:
https://github.com/openstack/horizon/blob/master/openstack_dashboard/local/local_settings.py.example#L35
is wrong - it says 'HTTP_X_FORWARDED_PROTOCOL'.
According to https://docs.djangoproject.com/en/1.7/ref/settings/#secure-
proxy-ssl-header (and our observations /
Public bug reported:
There is some javascript which hard codes the logout url to be
/auth/logout at:
https://github.com/openstack/horizon/blob/master/horizon/static/horizon/js/angular/horizon.js#L29
This is incorrect in number of distributions / deployments where horizon
resides in a path such
Public bug reported:
We have a user who is close to their quota, and trying to create a
snapshot that says
Error Unable to create snapshot
That confuses users. The real error is in the nova log.
OverLimit: VolumeSizeExceedsAvailableQuota: Requested volume or snapshot
exceeds allowed Gigabytes
Public bug reported:
parts of the code do this:
from oslo_config import cfg
other parts do this:
from oslo.config import cfg
There should be one version we use and the usage should be the same /
very similar.
** Affects: horizon
Importance: Undecided
Assignee: Nikunj Aggarwal
Public bug reported:
I have found issues where the angular cookies js file is not being
found, unless I include this in my settings file:
import xstatic.pkg.angular_cookies
...
STATICFILES_DIRS.append(('horizon/lib/angular',
Public bug reported:
We are having problems where the admin volumes page is bombing out /
error / http 500. This seems to occur due to an issue in
openstack_dashboard/dashboards/project/volumes/volumes/tables.py
in the function def get_attachment_name(request, attachment):
The problem seems
Public bug reported:
If a user has the admin role, and it working in a particular project -
some quota lookups look for items used globally, and do not make a
tenant scoped api call.
For example:
https://github.com/openstack/horizon/blob/master/openstack_dashboard/usage/quotas.py#L303
This api
Public bug reported:
When a user has selected region B, and then changes to a different
project, the region is selected back to region A (or the default/first
region).
This confuses users and puts the burden on them to really understand the
project and region scoping activity.
** Affects:
Public bug reported:
Using the plugin approach to add a panel to an existing dashboard, all
the dashboards become reordered. It's not clear how or why this occurs.
** Affects: horizon
Importance: Medium
Assignee: Eric Peterson (ericpeterson-l)
Status: Confirmed
** Changed
Public bug reported:
Steps to reproduce:
1) Create a group
2) Modify the group members
3) click the add button
(the add button fails to do anything)
The problem is somewhere around the _modal_form_add_members.html
file in this code:
/* as you navigate around the more pages, they stack
Public bug reported:
Using the access security screen, then creating a new keypair
Users are directed to a success page where the download of the keypair
starts automatically. However, there is also a download keypair link
which consistently fails.
It seems like maybe either we need to
is not
required).
** Affects: horizon
Importance: Undecided
Assignee: Eric Peterson (ericpeterson-l)
Status: New
** Changed in: horizon
Assignee: (unassigned) = Eric Peterson (ericpeterson-l)
--
You received this bug notification because you are a member of Yahoo
thats not true... the all_tenants=True flag ensures more instances are
retrieved.
https://github.com/openstack/horizon/blob/master/openstack_dashboard/dashboards/admin/instances/views.py#L71
** Changed in: horizon
Status: Invalid = Confirmed
--
You received this bug notification
62 matches
Mail list logo