Switches to using keystone sessions and adds compat between V2.0 and V3
identity APIs.
This additionally folds in a small change to fix doc generation and
unwedge the gate. This was originally proposed in another set[1].
[1] https://review.openstack.org/#/c/273887
Change-Id: Iaa34b620cb45cfb82f9294289a89c19c2b097919
This beefs up the functional test suite to do some tests on a
tenant router. The test can now creates one-off tenants to be used
in the tests. The new tests act entirely on behalf of the test
router and attempt on only use admin-level things when required.
Partially-Implements: blueprint ci-updates-mitaka
Change-Id: I26fa70f877522f09075dc87583f2359cc0dcaf41
This switches keystone to use oslo-config-generator, where the contents
of our sample configuration file are configured using a configuration
file in etc/oslo-config-generator.
Also fix some config problems.
blueprint autogen-astara-conf-file
Change-Id: I394805b18eecc4fbc583f9d64d34b8e95b55a845
In this step all of the imports and usage of akanda.rug is updated to
use astara. Addtionally rename all internal references from Akanda to
Astara.
Change-Id: I0cb8596066d949bceaadc4718b210fc373b5f296
Depends-On: I87106ae63747291bb6424839b5155f53136c54f9
Implements: blueprint convert-to-astara
This is the first step in converting the repo for Astara. This includes
changes to console scripts, new default users and packaging hooks.
Change-Id: If8e383e86cf867b8167f22b4411edf167d6cbd9a
Partially-Implements: blueprint convert-to-astara
This abstracts away how/where akanda.rug.api.nova.boot_instance()
creates a new resource. It preserves existing behavior of booting
one-off instances for resources in the OnDemandInstanceProvider.
It introduces a PezInstanceProvider which relies on a new service
akanda-pez-service. This service is responsible for keeping pools
of hot-standby nodes to be ready for use as appliance VMs. The size
of said pools are configurable and it maintains separate pools for
each enabled driver type.
When configured with the the PezInstanceProvider, the Rug will request
instances via RPC from Pez instead of booting directly from Nova.
To enable, you must run the akanda-pez-service and then set
'instance_provider=pez' in rug.ini. The default, which preserves existing
behavior, is 'instance_provider=on_demand'
Implements: blueprint nodepool-support
Change-Id: I7680c747d8c351c63d675f5fb377eee5edbf7b31
This abstracts all router-specific functionality into a router driver,
paving the way for the Rug managing other advanced services.
'rug-ctl router {debug, manage, update, rebuild}' commands have been
deprecated in favor of 'rug-ctl resource' equivalents, to be removed
in M. The newer 'resource' commands are meant to address any
Rug-managed resource. These requests are routed to the correct place
on the server-side, instead of a Neutron lookup on the client side.
Co-authored-by: Adam Gandelman <adamg@ubuntu.com>
Partially Implements: blueprint appliance-provisioning-driver
Change-Id: I1d2c68b81d28ebb0ca6ce4dcad237c4803bb96d6
In preparation for scale out RUG, this adds a database layer
(built on oslo.db) that will be used for managing debug modes.
Instead of tracking debug'd/ignored routers and tenants in-memory
or on disk as files, this uses a database. This means that putting
things into debug mode via rug-ctl are now persistent, and the file-based
approach is no longer available. A sqlite database (the default) can be
used for single node installs, or the RUG can be pointed at mysql/pg to
handle this in larger environments.
This also adds a global debug mode that can be used to ignore all events
during maintanence periods.
A new optional 'reason' argument has been added to the debug modes, allowing
operators to add a note when entering a tenant/router/cluster into debug
mode.
Change-Id: I3f5129e11b11cf5aaed8889da3b204104e5ad203
Closes-bug: #1470619
Partially implements: blueprint rug-scaling
Add a command to forcibly rebuild a router. It works
by destroying the instance and building a new one, which is
brute-force, but easy to understand.
I also changed the worker's lookup for tenant router
managers so the tenant id is normalized cleanly to
avoid confusion since some events use '-' and some
do not (yay, consistent!).
Change-Id: I79a622b0c6cc23041da55ec2f54f79ccf463379e
Add a command to have the worker reload the configuration file(s). This
will be useful when we automatically deploy new router images, since
we can change the running RUG without restarting it and losing any work
that is happening at the time.
Change-Id: I08628b1292e0e5a26dc3382c790dd41a806716bb
Andreas Jaeger