Updates to use new Shipyard pod configurations, removes node port use.
Requires Shipyard commit a11e962eef5a5aa8f8fc15c4a324dfa6b2465061 or
newer.
Change-Id: Iff5af45d6b4f60c34d18aad821cf9afa4ea67dda
Transposed characters in the target replacements in
some of the deployment files can lead to unexpected
results (mostly that defaults are used instead of
overrides)
Since defaults were being used before, the endpoint definition needed to
be redefined to use the expected scheme (amqp) and path (/airflow)
Change-Id: I85a58c7b525e29208957845bf25ffbb0111c8f35
* "keystone-api" host is used for internal communication and
it listens on 5000 port.
* "keystone" host is used for public and admin, its port
is 443.
Ensure default host and port match if "host_fqdn_override"
is not specified.
This is also consistent with OSH defaults.
Change-Id: I020c5a3c13ea33849cd794b845427fbe03e914c5
- Due to latest changes client port is no longer exposed directly,
and charts using elastic should use port 80.
This fix relaxes port overrides to allow using chart defaults.
Change-Id: I8dac2d7240f3e8e925340b2479034a059ab85d42
This updates chart configuration overrides for the monitoring
services, as well as adds missing secrets, charts and service
accounts for other exporters that have now been enabled
Change-Id: Ia1ed7bba38d7c262e85de8162d53012cdadf487e
This updates the Elasticsearch and Fluent-logging charts to use
the most recent configuration keys in their values overrides, and
also introduces support for the ceph-rgw s3 api for use for
Elasticsearch snapshot repositories
Change-Id: Ia998db9006350a22fcc7dc3052301d7a5b8259f4
- Add tenant ceph manifests in globals
- Add reference overrides for site in airship-seaworthy
- Re-purpose 4th control as an additional compute to total 3.
This allows to maintain proper ceph replication
- Update disk layout to use partitions for journals
Change-Id: I0edf35bd56e4e5253f78c9074d88369283b6fac3
Notable changes
- Ceph client split in provisioner and rgw charts
- Fixes in endpoints/osh-infra
- Multiple charts moved from openstack-helm to openstack-helm-infra
- Hold back MaaS/Drydock helm-toolkit - fixes in charts needed
Change-Id: I790865a4e1ec973b8182d66acad681b99414187c
- Set correct release group prefix
- Remove DNS and HTTPS from endpoints
- Fix external interface for ovs/neutron
- Enable all charts in full-site
- Set versions to known working state
Change-Id: I5ed4cecc3f0034da448d10b79ccb0abf4efc2b86
- Fix missing k8s IP for etcd certs
- Fix release groups for wait labels
- Fix versions that changed since initial manfiests
- Fix barbican override to use loci image
- Add re-generated certs
- Adjust partitions for host profiles
Change-Id: I961886cbaa3a0ce03e836d2392cf81bb74afc26c