python-tripleoclient provides the following openstack baremetal
commands:
openstack baremetal instackenv validate:
openstack baremetal import:
openstack baremetal introspection bulk start:
openstack baremetal introspection bulk status:
openstack baremetal configure ready state:
openstack baremetal configure boot:
Based on the data below several of these are already deprecated and have
been so longer enough to be removed.
openstack baremetal instackenv validate:
tripleoclient.v1.baremetal:ValidateInstackEnv
NOT Deprecated
openstack baremetal import:
tripleoclient.v1.baremetal:ImportBaremetal
DEPRECATED in b272a5c6 2017-01-03
New command: openstack overcloud node import
openstack baremetal introspection bulk start:
tripleoclient.v1.baremetal:StartBaremetalIntrospectionBulk
DEPRECATED in b272a5c6 2017-01-03
New command: openstack overcloud node introspect
openstack baremetal introspection bulk status:
tripleoclient.v1.baremetal:StatusBaremetalIntrospectionBulk
NOT Deprecated
openstack baremetal configure ready state:
tripleoclient.v1.baremetal:ConfigureReadyState
NOT Deprecated
openstack baremetal configure boot:
tripleoclient.v1.baremetal:ConfigureBaremetalBoot
DEPRECATED in b272a5c6 2017-01-03
New command: openstack overcloud node configure
This leaves:
openstack baremetal instackenv validate
- This is somewhat superceded by the mistral validation in
tripleo-common
openstack baremetal introspection bulk status
- This should have been deprecated along with
'openstack baremetal introspection bulk start' and isn't useful without
the former.
openstack baremetal configure ready state
- Seems to only support drac and requires a datafile no loner generated
by tools
As these commands have outlived their useful lifetime and do not require
deprecation we're free to remove them so let do it.
Change-Id: Ie6b6a8578e4d12503a3dbfa5747309033d53466e
The default file is now v3, so these will always be the same.
Related-Bug: #1733640
Depends-On: I5e779f48a02c5389cebf84e9d3899250d63cace8
Change-Id: I3612c58c356f8955bd44655cf5aacc20be532cbc
The tripleoupstream registry is deprecated in favor of tripleomaster,
where images are pushed automatically on rdo promotion.
Change-Id: I872a04f6c7d1c256d126b32303d73bafc6e8fd89
This setting doesn't exist in instack-undercloud and it has
long since been removed from t-h-t as well. There is no longer
any use for it.
Change-Id: I4aec82225c056e8cc7bf96a53bc9ce95957f9a2c
- SERVICE_LIST is not used anywhere, let's drop it.
- Same for _get_password, _get_base_service_data and
_get_endpoint_data
Change-Id: Iff55d3b83ac4ad042a5b596c5b435877b588fe5c
The default value for --overcloud-ssh-key was wrong (it was actually a
username). Instead of using a default value at all, just remove the
default since we can't sanely pick one.
Change-Id: I57a7293f0cc08b0fab0ee7c2f0b4c0a8b2bd2f44
Closes-Bug: #1736606
Ironic is going to deprecate classic drivers in Queens and remove them in Rocky.
This change enables hardware types ilo and idrac that correspond to alreadys
enabled classic drivers pxe_ilo and pxe_drac. It also adds support for others
common hardware types, but does not enable them by default.
The enabled_drivers option is deprecated.
This change mostly copies commit 58701fe014bf118ee0bbed5272c1a08e3f594ae0
from instack-undercloud, but adapts it for heat-driver undercloud.
Also adds staging-ovirt from review I3e5d0a72e2ba22537ce2c0eb30000c29f429dc3f
Change-Id: I3c20591a9900afb60aadcfa151b9ff29e7d3e16d
We do not need tox_install.sh, pip can handle constraints itself
and install the project correctly. Thus update tox.ini and remove
the now obsolete tools/tox_install.sh file.
This follows https://review.openstack.org/#/c/508061 to remove
tools/tox_install.sh.
Change-Id: If0b9a678354b634a2f060df4a6ff1b479967d566
For Nova+Ironic deployments we could run the enable_ssh_admin Mistral
workflow directly and take advantage of the public key injection via
os-collect-config, but since the script seems to work for both
deployed-server and Nova/Ironic envs, we can just use it
unconditionally and keep the logic DRY.
Change-Id: I2bae54d282591abc8a014cdce805e90305a7a34d
Closes-Bug: #1735173
On minor update, when the execution is still running
but there is no claim message anymore, we should
capture the zaqar exception to not print it to the
console
Change-Id: I802ffd553c54e4a4f9998420645aa078f508b9f0
Closes-Bug: #1734957
The qemu-img-ev package comes from the virtualization sig backports.
These stopped getting installed with
I747c2b8754effbc6ec82af3bf7543fd9599a6c14. I believe that this is
really installing these dependencies for dib to convert the image, and
that works with the standard qemu-img package.
Change-Id: I0c09c54f7ace4ac28942b733cbc6dfe42436ecdc
We noticed that heat was sleeping a lot waiting on disk. By putting the sqlite
database on a ramdisk (tmpfs) we cut the deploy time in half.
Change-Id: Ic22bf49787279993d5d6ee081b31d789b519e555
There were two issues: The action timeout and the wait_for_messages
timeout.
The wait_for_messages did not need a timeout here and it was causing the
bug. I added an cli option for the timeout and setup a reasonable
default (the previous default was over 4 hours that would invalidate
keystone session duration)
Change-Id: I232d071516d3f53f9f87143d9d65ebe3e58b0f38
Closes-Bug: 1734712
Roles which don't have any instances will not have a 'ctlplane' record
in the RoleNetHostnameMap, so we shouldn't be depending on it being
always present.
Change-Id: I2da025961c584cb1adc83943561b1d9faa3559b1
Closes-Bug: #1734703
The rolling-update for ceph ansible is broken now,
so we have to switch to the docker site playbook
instead
Closes-Bug: 1734307
Change-Id: I41f638b0a7df903bd95ab8d42241249d6ec1043a
Allow to parametrize the hardcoded 'undercloud' domain name.
F.e. RDOcloud uses an 'rdocloud' domain, therefore some DNS things might be
failing with the hardcoded value.
The default domain name remains unchanged. It can be
overriden with a new '--local-domain' option as well.
Related-bug: #1698349
Partial-bug: #1733873
Change-Id: Iad63142b241f9f7f0bbef68473f18b69f36daf76
Signed-off-by: Bogdan Dobrelya <bdobreli@redhat.com>
The 'fake' file was being created and left on disk when running the
unit tests, sometimes causing random failures during directory
creation. This mocks the function in all tests to avoid this, and adds
a new unit test specific to it to ensure we don't lose coverage.
Change-Id: I6cf6ec5e85af4a748e00383f7ee65636eb5a3251
Closes-Bug: #1732010
Consume the update zaqar queue to get the ansible output in
real time.
The queue used is not the 'tripleo' queue because we need
to not be disrupt by other messages.
We need to claim all the zaqar messages in the queue to
get a consistent ansible output
Change-Id: I3682051eb719f8da9c744f7ec8be3a58f3db3f86
Closes-Bug: 1732497
