1) UCP -> Airship
2) readthedocs.org -> readthedocs.io (there is redirect)
3) http -> https
4) attcomdev -> airshipit (repo on quay.io)
5) att-comdev -> openstack/airship-* (repo on github/openstack git)
6) many URLs have been verified and adjusted to be current
7) no need for 'en/latest/' path in URL of the RTD
8) added more info to some setup.cfg and setup.py files
9) ucp-integration docs are now in airship-in-a-bottle
10) various other minor fixes
Change-Id: I4b8cc6ddf491e35d600a83f5f82d7717108e31dd
We will move away from installation of shipyard CLI/client
binaries and will make use of the shipyard docker container
instead for our CI/CD integration environment.
Change-Id: I253cbc1e2f9310b9c11195e1c68b97ee8d0efead
We will no longer need to manually update the '/etc/hosts' file
for 'in-cluster' scenario with the bug fix in [0]. This patch set
is meant to update the scripts in the tools directory to reflect
that change.
Note that user will still need to ensure that the DNS is set up or
'/etc/hosts' file is properly updated if he/she is executing the
workflow from outside the cluster, e.g. via jump server
[0] https://review.gerrithub.io/#/c/398739/
Change-Id: I0580cfc473b0f1b0ed824a578d3e6c89dd6f1443
This Patch Set is meant to address Shipyard issue #18 [0]
Issues Identified:
1) The deckhand_load_yaml.sh script does not check if the
ingress_controller_ip returned a null value. This may
result in invalid entries written to /etc/hosts.
2) The subsequent DNS entries written to /etc/hosts are not
idempotent, so script re-runs (e.g., after discovering and
resolving YAML validation errors) will result in duplicate
(and possibly disagreeing) DNS entries written to /etc/hosts.
3) The '--replace' flag should also be used in the "shipyard create configdocs"
command, to allow for iterative YAML corrections and re-execution
of this script to get past validation errors.
[0] https://github.com/att-comdev/shipyard/issues/18
Change-Id: I33531d07a5d6c425071af99c8824c12d3e177e69
[0] allows namespace specific ingress controllers and a
cluster wide ingress controller that acts as the external
front door to the cluster.
As the codes in OSH have been merged, we are removing the
workaround that was previously put in place for UCP namespace.
[0] https://review.openstack.org/#/c/533271/
Change-Id: Ia1c3e85e55678d62a4eb8506841dffc547012ef0
We are getting the following errors in [0] when executing
the shipyard CLI due to timeout issue.
We will be increasing the timeout from 30 seconds to 60
seconds to allow more time for site validations.
There is another problem with the deckhand_load_yaml script
as the environment variables were defined but not exported.
This p.s. is meant to address these 2 issues.
[0] Error messages
root@genesis:~# shipyard commit configdocs
Error: Validations failed
Reason: Validation
- Error: Drydock unable to validate configdocs
- Error: HTTPConnectionPool(host='drydock-api.ucp.svc.cluster.local', port=9000): Read timed out. (read timeout=30)
- Error: Armada unable to validate configdocs
- Error: HTTPConnectionPool(host='armada-api.ucp.svc.cluster.local', port=8000): Read timed out. (read timeout=30)
root@genesis:~#
Change-Id: Ia04d244077c9f809e19aca1b47ce178eb5984b0c
We will make use of this script to load the required YAMLs
into Deckhand using Shipyard CLI.
We will move the Shipyard CLI installation steps and updating
of /etc/hosts from the 'deploy_site' script to 'deckhand_load_yaml'
script.
Note that we will need to execute the 'deckhand_load_yaml' script
first before running the 'deploy_site' script
Change-Id: If463f606c3c7e99763c0d8852f729a9a3e270425
Dustin Specker