Gate scripts are trying to calculate the SUBNET_RANGE basing on IP address
of default interface. There is ipcalc tool used.
The issue is that based on infra datacenter selected to run the VM gate,
the default interface can have address with /32 network.
If IP has /32 network, the ipcalc would not calculate the network address,
for example 192.168.0.0/24, which is needed for ceph to be configured.
As a workaround, try to change the /32 to /24 after getting the IP of the
default interface, and then calculate the network address.
Change-Id: Ia2858c6a2b4c56fbc42ef06b5f157625a56beb3c
This PS update the logic used to determine the default tunnel
interface if one is not specified, as the original method could fail with
complex device names.
Change-Id: Ifec81f517ee29bef0926779591ab9cf5fe0e63c4
New kubernetes-entrypoint version was released. K8s-entrypoint
authors maintain images at Quay. The image uses CoreOS, which
is more lightweight than the current Ubuntu image, so it
should lessen the burden on the infrastructure.
Change-Id: Id8c2a4d065550ffbd64476377247cccf213b58e1
Partial-Implements: blueprint entrypoint-namespaces
Kubernetes 1.8 is stricter about the feilds in a secrets manifest,
this PS updates OpenStack-Helm to be compliant.
Change-Id: I9e19d07060d8517e0f4fd3056013191b1b4ba2da
When there are any defunct charts either because they were removed or
perhaps existed when testing another branch (commonly happens when
testing with git review -d ...) the build process breaks.
This should remove them.
Change-Id: I850d6ad820148f657f253bfdf1a765c673af7099
This devref is covering the existing OpenStack-Helm neutron
architecture and proposal for introducing multiple SDNs.
Added linuxbridge usage docs.
Change-Id: If6c8753d85ea2bd2fd7df77e84b5e61bf64dc507
Implements: blueprint support-linux-bridge-on-neutron
Co-Authored-By: Gage Hugo <gagehugo@gmail.com>
This PS moves the rally test config into values.
Partially implements: blueprint remove-pregenerated-config-templates
Change-Id: Ic051e762d792308de1d6038c37ca71c3efc9698c
This PS adds a ceilometer chart.
Change-Id: I8ddf3867514f2bbf0d9d0607733594a512c75e35
Co-Authored-By: Steve Wilkerson <sw5822@att.com>
Co-Authored-By: Larry Rensing <lr699s@att.com>
With auto_bridge_add defined as:
auto_bridge_add:
# nothing here
helm doesn't treat is as am empty table so issues warnings such as:
2017/10/03 19:24:37 warning: destination for auto_bridge_add is a table. Ignoring non-table value <nil>
This happens most of the time as we use `helm --values=...` to augment
values.yaml in the chart.
This change clarifies it is indeed a table and the warnings no long
appear.
Change-Id: Ib733ed72df49e134a9199e4024b2893629b990dc
This PS adds a gnocchi chart.
Change-Id: I51634a3876a36dde1eb28cc62cef3189e345cd2c
Co-Authored-By: Steve Wilkerson <sw5822@att.com>
Co-Authored-By: Larry Rensing <lr699s@att.com>
In addition to restoring the secret setting again this removes some
redundant logic and does minimal cleanups.
Change-Id: I5dbcbd393b18befd762518e3865da9e006cf5bd9
This PS adds a PostgreSQL chart.
Change-Id: Id33b110e242d86a78215bc3926bfba4602b9a1c9
Co-Authored-By: Steve Wilkerson <sw5822@att.com>
Co-Authored-By: Larry Rensing <lr699s@att.com>
The movement of the ceph secret logic into the start script introduced
a bug that was missed in review when ceph is disabled, this PS
restores expected operation.
Closes-Bug: #1720061
Change-Id: Ie373ff94814ea396832a83101c885ebf92b6f8cc
This PS moves the neutron credentials from the config section
of the values.yaml to the endpoints section, seperating them from
the service configuration.
Change-Id: I548148449f56e4eea92669ad8e98e3eff1ca7b27
This PS moves the nova credentials from the config section
of the values.yaml to the endpoints section, seperating them from
the service configuration.
Change-Id: Idfc696e0fc907b702a8ca81d81e50cb252cb7af8
Implemented Kubernetes jobs to drop a database for an OpenStack service
on deletion of Helm chart. This job is configurable to be enabled or
disabled in the values.yaml file for each chart.
Change-Id: Ia1a0834cb43c7e883aaf507a7c7683fa1cf4e838
In Kubernetes version 1.7.x, /proc/1/fd/2 command in rabbitmq readiness
shell script makes an error. Instead of using /proc/1/fd/2, use /dev/null.
Change-Id: Ia4fb7f3130ec1a8fccc125bf11e997b77935e1ba
Closes-bug: 1719763
The new storage-init script doesn't handle the case in which the rbd pool
user already exists. This PS solves the issue by improving the script.
Change-Id: I22c55429d3077393b279fb784845db41a69b5469
Closes-Bug: 1719081
