horizon/openstack_dashboard/usage/views.py

205 lines
7.3 KiB
Python

# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
import collections
from django.contrib.humanize.templatetags import humanize as humanize_filters
from django.utils.translation import pgettext_lazy
from django.utils.translation import ugettext_lazy as _
from horizon import exceptions
from horizon import tables
from horizon.templatetags import sizeformat
from openstack_dashboard import api
from openstack_dashboard.usage import base
class UsageView(tables.DataTableView):
usage_class = None
show_deleted = True
csv_template_name = None
page_title = _("Overview")
def __init__(self, *args, **kwargs):
super(UsageView, self).__init__(*args, **kwargs)
if not issubclass(self.usage_class, base.BaseUsage):
raise AttributeError("You must specify a usage_class attribute "
"which is a subclass of BaseUsage.")
def get_template_names(self):
if self.request.GET.get('format', 'html') == 'csv':
return (self.csv_template_name or
".".join((self.template_name.rsplit('.', 1)[0], 'csv')))
return self.template_name
def get_content_type(self):
if self.request.GET.get('format', 'html') == 'csv':
return "text/csv"
return "text/html"
def get_data(self):
try:
project_id = self.kwargs.get('project_id',
self.request.user.tenant_id)
self.usage = self.usage_class(self.request, project_id)
self.usage.summarize(*self.usage.get_date_range())
self.kwargs['usage'] = self.usage
return self.usage.usage_list
except Exception:
exceptions.handle(self.request,
_('Unable to retrieve usage information.'))
return []
def get_context_data(self, **kwargs):
context = super(UsageView, self).get_context_data(**kwargs)
context['table'].kwargs['usage'] = self.usage
context['form'] = self.usage.form
context['usage'] = self.usage
try:
context['simple_tenant_usage_enabled'] = \
api.nova.extension_supported('SimpleTenantUsage', self.request)
except Exception:
context['simple_tenant_usage_enabled'] = True
return context
def render_to_response(self, context, **response_kwargs):
if self.request.GET.get('format', 'html') == 'csv':
render_class = self.csv_response_class
response_kwargs.setdefault("filename", "usage.csv")
else:
render_class = self.response_class
context = self.render_context_with_title(context)
resp = render_class(request=self.request,
template=self.get_template_names(),
context=context,
content_type=self.get_content_type(),
**response_kwargs)
return resp
ChartDef = collections.namedtuple(
'ChartDef',
('quota_key', 'label', 'used_phrase', 'filters'))
# Each ChartDef should contains the following fields:
# - quota key:
# The key must be included in a response of tenant_quota_usages().
# - Human Readable Name:
# - text to display when describing the quota.
# If None is specified, the default value 'Used' will be used.
# - filters to be applied to the value
# If None is specified, the default filter 'intcomma' will be applied.
# if you want to apply no filters, specify an empty tuple or list.
CHART_DEFS = [
{
'title': _("Compute"),
'charts': [
ChartDef("instances", _("Instances"), None, None),
ChartDef("cores", _("VCPUs"), None, None),
ChartDef("ram", _("RAM"), None, (sizeformat.mb_float_format,)),
]
},
{
'title': _("Volume"),
'charts': [
ChartDef("volumes", _("Volumes"), None, None),
ChartDef("snapshots", _("Volume Snapshots"), None, None),
ChartDef("gigabytes", _("Volume Storage"), None,
(sizeformat.diskgbformat,)),
]
},
{
'title': _("Network"),
'charts': [
ChartDef("floatingip", _("Floating IPs"),
pgettext_lazy('Label in the limit summary', "Allocated"),
None),
ChartDef("security_group", _("Security Groups"), None, None),
ChartDef("security_group_rule", _("Security Group Rules"),
None, None),
ChartDef("network", _("Networks"), None, None),
ChartDef("port", _("Ports"), None, None),
ChartDef("router", _("Routers"), None, None),
]
},
]
def _apply_filters(value, filters):
if not filters:
return value
for f in filters:
value = f(value)
return value
class ProjectUsageView(UsageView):
def _get_charts_data(self):
chart_sections = []
for section in CHART_DEFS:
chart_data = self._process_chart_section(section['charts'])
chart_sections.append({
'title': section['title'],
'charts': chart_data
})
return chart_sections
def _process_chart_section(self, chart_defs):
charts = []
for t in chart_defs:
if t.quota_key not in self.usage.limits:
continue
key = t.quota_key
used = self.usage.limits[key]['used']
quota = self.usage.limits[key]['quota']
text = t.used_phrase
if text is None:
text = pgettext_lazy('Label in the limit summary', 'Used')
filters = t.filters
if filters is None:
filters = (humanize_filters.intcomma,)
used_display = _apply_filters(used, filters)
# When quota is float('inf'), we don't show quota
# so filtering is unnecessary.
quota_display = None
if quota != float('inf'):
quota_display = _apply_filters(quota, filters)
else:
quota_display = quota
charts.append({
'type': key,
'name': t.label,
'used': used,
'quota': quota,
'used_display': used_display,
'quota_display': quota_display,
'text': text
})
return charts
def get_context_data(self, **kwargs):
context = super(ProjectUsageView, self).get_context_data(**kwargs)
context['charts'] = self._get_charts_data()
return context
def get_data(self):
data = super(ProjectUsageView, self).get_data()
try:
self.usage.get_limits()
except Exception:
exceptions.handle(self.request,
_('Unable to retrieve limits information.'))
return data