Since most services use OPENSTACK_API_VERSIONS for the major
API version only but Magnum uses it for micro-versions, it ends
up with some decimal values.
At the moment, we have it set to 1.10 however since it's being
set as a number, it's becoming 1.1 which is too old. The
solution is to quote it as a string instead and that is what
magnum-ui expects as well.
Change-Id: I46b5fec69d8bc353e38a6f8e6e506e2a460371f4
This change updates all Ceph image references to use Focal images
for all charts in openstack-helm.
Change-Id: I67cd294e2aabf3c3af404da42204f9b6157b06f7
Beginning with the Pacific release, Ceph pools are not allowed to
use 1x replication by default. This is problematic for the
openstack-helm gate scripts, which frequently use 1x replication
for automated testing. This change adds Ceph configuration and
command overrides to allow those gate scripts to continue to use
1x replication for testing.
Change-Id: I21ed3e43f3773d5ea830959f1b66b35f38185ca7
This PS is to add a timeout to the readiness probe. This timeout
is needed in certain environments where the default value of 1
is insufficient.
Change-Id: I0acb6438be9de150b4c36da537379e6032d9646f
This change adds the endpoint details for Nova to allow for online
resizes however a very key thing in this is that it is actually
using the Cinder credentials to talk to Nova.
OpenStack projects have historically arbitrarily decided to use
the user credentials of the _target_ service rather than the
source service which does not make sense, a mailing list discussion
seemed to have not brought up any negatives but only positives.
For the future, we can continue to do this which will simplify
our endpoints section but this is a start.
Change-Id: Ib9b500ef9a9bc34c8b64215bee57451494735573
We have a few deprecated config options that are not being
used anymore as well as some that have been moved to other
groups for quite sometime.
Change-Id: Ibd447897f6399bab47b031ccab228188ebed8266
bugfix: add helm3_hook default
bugfix: revert change on neutron - not applicable to this changeset
Change-Id: I4bc60e8e34a6861742f1f9e7582e69f49740ab87
OpenStack services already moved to use policy in code.
No need to have policy file at this point, at least no need to put
default policy rule to policy.yaml file anymore.
To put in duplicate rules, will cause unnecessay logs and process.
Also not healthy for policy in code maintain as the `default` rules in
openstack-helm might override actual default rules in code which we
might not even mean to change it at all.
Change-Id: I29ea57aa80444ed64673818e597c9ca346ba7b2f
Change ceph-config-helper image tag to bionic from focal. There is no
focal based image available.
Change-Id: I6c4039161d09a55cac84fc047080603953a1deb5
This PS adds backoffLimit to nova-bootstrap job in nova chart. By default, this job was created from a template in helm-toolkit.
58291db1a6
In this commit the job was re-designed without controlling of the backoffLimit value.
Change-Id: Icb28363be8063d849fd22e9c2542edf1eb203d60
We dropped train support a long time ago now, and our latest efforts
are to drop ussuri/bionic images. This change removes any leftover
train overrides as well as any ussuri overrides. This also changes
any image defaults to use wallaby.
Change-Id: I818a3a79faa631ec1b7de625f2113c6f19610760
When the placement chart was originally written, there was a
need to migrate from an existing nova-placement setup.
Now that nova and placement have been separated for several
releases, there's not much use in maintaining this job. This
change removes the db-migrate functionality in the placement
chart due to issues with it in newer releases.
The chart version has been bumped to 0.3.0 to signify a
non-trivial change in functionality as well.
Change-Id: I6ff802ab8356deb5e927f414500c52b663cfa30b
The list-agents rally test for nova was removed in wallaby, but it
was also only supported by the XenAPI hypervisor driver. We have
specifically overriden it for newer releases of openstack, but
with its removal and the specific driver usage, there's no real
need to keep it around.
Change-Id: I056b397444e8dc5d4b256a6fe03c23b53a0c0fff
This reverts commit d554d74bf0c6d61d63f92e1d8224881dd4fa726d.
This change is very problematic, because then the volumes will no longer be able to have API operations work on them, since the volumes are tied to a host, so when that host it down, things are bad.
We should keep this old (non-clean behaviour) and switch towards using Cinder clusters instead https://docs.openstack.org/cinder/latest/contributor/high_availability.html
This means configuring a few other things, but this is a really bad change and will break production deployments. I suggest we revert ASAP.
Change-Id: Iee54ef5feca5c8bee80aded75d2fd182a431adb3
Now `cinder-volume-worker` is set as DEFAULT.host. It results only
one host per volume service type registered regardless of replica
numbers.
To fix this, removed DEFAULT.host from default values so hostname
of pods will be used as host name.
It will produce garbage services with down state because pod name
changes every creation. To avoid this, added a cronjob for service
clean.
Change-Id: I9ec6f62105124f9088afdb231099b532fc83bb34
