Age | Commit message (Collapse) | Author | Files | Lines |
|
The DIB variables only make sense on bifrost so they should live in the
bifrost provision script.
Change-Id: I05baca7b9c6889a0d401a9288998991324d270af
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
The XCI deployment script may wipe the PXE directories so we need
to place the OS images after this has happened so we don't download them
every time.
Change-Id: Iccf943bc75592e7f5dddea59dfee14b2b080e7ef
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
We already have images for all 3 distros which we could simply re-use
for all the XCI VM deployments instead of building new ones everytime
with DIB. The images will be copied to the new VM from the cache
directory if they are available otherwise we will simply download them
during the XCI execution phase.
Change-Id: I2a8391650558511668654c6b54a10db316f867a2
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
The lack of UC means that pip is attempting to install libvirt-python 4.1.0,
which is failing on Ubuntu due to
https://bugs.launchpad.net/openstack-requirements/+bug/1753539.
Currently upper-constraints limits libvirt-python to 4.0.0, which does not
have this issue.
Change-Id: Idc97a0d9928cebafc9167614723236d2584cc93f
Signed-off-by: Fatih Degirmenci <fdegir@gmail.com>
|
|
From upstream docs:
'The host-model mode is essentially a shortcut to copying host CPU
definition from capabilities XML into domain XML. Since the CPU
definition is copied just before starting a domain, exactly the
same XML can be used on different hosts while still providing
the best guest CPU each host supports'
It's probably safer to use that instead of 'host-passthrough' so
we can get a CPU (both for main VM and the nested ones) that libvirt
understands. Moreover, it's important to present a CPU that the
guest OS understands as well.
Change-Id: I25a8ff0e8635df9804c793d184f048cc86059ce0
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
According to the docs[1]
"writeback: This mode causes the hypervisor to interact with the disk
image file or block device with neither O_DSYNC nor O_DIRECT semantics.
The host page cache is used and writes are reported to the guest as
completed when they are placed in the host page cache. The normal page
cache management will handle commitment to the storage device.
Additionally, the guest's virtual storage adapter is informed of the
writeback cache, so the guest would be expected to send down flush
commands as needed to manage data integrity. Analogous to a raid
controller with RAM cache."
and
"writeback: This mode informs the guest of the presence of a write
cache, and relies on the guest to send flush commands as needed to
maintain data integrity within its disk image. This is a common
storage design which is completely accounted for within modern file
systems. This mode exposes the guest to data loss in the unlikely case
of a host failure, because there is a window of time between the time
a write is reported as completed, and that write being committed to the
storage device."
"unsafe: This mode is similar to writeback caching except for the
following: the guest flush commands are ignored, nullifying the data
integrity control of these flush commands, and resulting in a higher
risk of data loss because of host failure. The name “unsafe” should
serve as a warning that there is a much higher potential for data
loss because of a host failure than with the other modes. As the
guest terminates, the cached data is flushed at that time."
It's beneficial to use the host page cache to cache I/O from the guest
instead of waiting for data to reach the actual disk device. We do not
normally care about data integrity so data loss is not a problem.
Moreover, we drop the cache configuration from the flavor files since
it's independent of the flavor that's being deployed.
[1] https://www.suse.com/documentation/sles-12/singlehtml/book_virt/book_virt.html#cha.cachemodes
Change-Id: I118ffdf84b1be672185b3eff60fe5d0b5f1a590d
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
When running the bifrost jobs for the OPNFV 3rd Party CI, the Ironic
variables are not defined and we get the following error:
./scripts/bifrost-provision.sh: line 112: BIFROST_IRONIC_INSPECTOR_VERSION: unbound variable
As such, we need to provide reasonable defaults when we are running this
script outside of XCI.
Change-Id: I9cc3ae1b70494637a85558b9cae99fc1eeaebc74
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
We should pin the SHAs for Ironic components as part of the regular
bifrost bump to ensure maximum long term stability. This also bumps
the bifrost SHA to the latest available one.
Change-Id: Ia66d8ab6566db58e2ddab5134eb8445bd84d2a8f
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
The Ironic components are currently being installed from PIP but we
should install them from the git repo instead so we can always try
the latest code from master or the pinned version when such
functionality is implemented.
Change-Id: I0138c54befe52ed9df8bf617b78e063df3cf8750
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
|
|
The default CPU model of QEMU GenuineIntel/QEMU Virtual CPU is not the
best option for vCPU in terms of performance. As such, lets use the
real host CPU. This has also been submitted upstream and the template
should be removed once the upstream patch has been merged and the SHA
for bifrost is bumped.
Change-Id: I23d86933604c290d5650b73f420972d0a2b23dc8
Link: https://review.openstack.org/#/c/528677/
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
These options seem to help with the functest execution
within XCI VM.
Change-Id: I3b9b6b02fc571feb42543025f8ab179c6db18e0a
Signed-off-by: Fatih Degirmenci <fdegir@gmail.com>
|
|
There is too much noise when XCI is cloning repositories, installing
packages etc so lets make the console output somewhat more readable.
Change-Id: I15667e0349f943ad5488daf0e3cea9336c9166d1
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
|
|
The XCI_ANSIBLE_VERBOSITY is normally exported by the XCI bootstrapping
process. However, sometimes it's necessary to run bifrost-provision.sh
by itself so we can debug machine provisioning. As a result of which,
we need to initialize this variable properly so we don't fail with
'unbound variable' errors.
Change-Id: I81c59cf8eb3369c001e952c2c0a2846ed8d95cdf
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
The latest bifrost code contains a fix when setting up DNS from a glean
configuration drive and also a few virtualenv improvements. This allows
us to get rid of all the workarounds we had in the past to switch
Ansible versions in the middle of the XCI deployment.
Change-Id: If49e290315ec96efdc07d04ff6624439c53aee19
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
The ANSIBLE_VERBOSITY env variable is actually used internally by
Ansible and exporting it actually makes Ansible super verbose even if
the variable is empty. This feature is going away in the future so we
introduce our own XCI_ANSIBLE_VERBOSITY variable which is simply passed
to the command line to control verbosity in a proper and deterministic
way.
Change-Id: I8a32eeb1a86b4cb0a9872a51bbe9f9624d096e39
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
The bifrost-provision.sh script will provision machines using Ansible
playbooks. If we run this script in an Ansible playbook, that means that
we will run Ansible with Ansible and this can only lead to further
confusion when debugging issues. As such, since we already have a script
to provision machines, lets use it directly from the xci-deploy.sh
script. This also reverts 3f04e1fd72b14420788af64b14c35a5f74727f82 which
added virtualenv support in bifrost since that causes more problems than
it fixes for the time being.
Change-Id: Id82b7c06a2af28b66f64f1966227888227ec276f
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
This is already done a few lines above since
I94874d600a29247b7826324cc111901995df90dc so there is no need to do it
twice.
Change-Id: I661516d9aadd265bf6869005718b2f51c57ff84d
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
Now that bifrost uses virtual env for Ansible we can go ahead and
bump the Ansible version to match the one from OSA so we can make use of
the latest features. Furthremore, we stop passing ANSIBLE_VERBOSITY to
the command line since it's an environment variable so Ansible already
knows about it.
Change-Id: I52d9a211fc0ced1f7830f6945b2943f35d38109b
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
bifrost forces us to use an old Ansible version which means we can't
take advantage of new features in XCI playbooks. As such, we move the
entire bifrost step to a virtual environment so we can have a newer
version of Ansible on the host.
Change-Id: I94874d600a29247b7826324cc111901995df90dc
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
We are seeing issues with selinux enabled from time to time. Having
selinux enabled has proven to be rather unstable and Centos7 not being
a gate on upstream bifrost does not help in fully testing this scenario.
As such, the best we can do right now is to disable selinux until Centos7
becomes a gate upstream.
Change-Id: I93a3414cfebc3c3ef4ac063c8e623f36f615455f
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
Switch default target OS to Ubuntu Xenial (16.04) which is also
the version which is supported by OpenStack Ansible.
Change-Id: I73adc6472b0fb572177e1508cf831b4d3190245c
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
The os_client_config Ansible module used by all the ironic-* roles
depends on the keystone roles so we need to make use of them even
though they do not affect the end result. This fixes the following
OPNFV CI problem due to not having a clouds.yaml file present to
be consumed by the os_client_config Ansible module.
fatal: [controller00]: FAILED! => {"censored": "the output has been hidden due to the fact that 'no_log: true' was specified for this result"}
An exception occurred during task execution. The full traceback is:
Traceback (most recent call last):
File "/tmp/ansible_ApkCUQ/ansible_module_os_client_config.py", line 75, in <module>
main()
File "/tmp/ansible_ApkCUQ/ansible_module_os_client_config.py", line 63, in main
for cloud in config.get_all_clouds():
File "/usr/lib/python2.7/site-packages/os_client_config/config.py", line 798, in get_all_clouds
cloud, region_name=region['name']))
File "/usr/lib/python2.7/site-packages/os_client_config/config.py", line 1071, in get_one_cloud
auth_plugin = loader.load_from_options(**config['auth'])
File "/usr/lib/python2.7/site-packages/keystoneauth1/loading/base.py", line 162, in load_from_options
raise exceptions.MissingRequiredOptions(missing_required)
keystoneauth1.exceptions.auth_plugins.MissingRequiredOptions: Auth plugin requires parameters which were not given: auth_url
fatal: [opnfv]: FAILED! => {"censored": "the output has been hidden due to the fact that 'no_log: true' was specified for this result"}
Moreover, we cleanup the ~/openstack directory which may contain some
bifrost artifacts such as a clouds.yaml file.
Finally, we use 'sudo -H' because for the keystone roles we need the HOME
variable to be set properly.
Change-Id: I45b08bd33dd8ea9505fe10eb4b2b10956b3b683c
|
|
This change
- fixes GIT_BASE in jobs
- adjusts build blockers
- fixes where the flavor vars should be sourced from
- sources flavor vars for bifrost periodic jobs
- renames the xcimaster node to opnfv to be consistent
- removes obsolete xci-provision.sh script
- enable additional openstack services including tempest to
ensure the deployment is sane
Change-Id: Ifdce1da52d68a26c2b87e1bd3b1996ec119d8d90
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
Daily jobs will use pinned versions of bifrost and openstack-ansible in
order to make sure we use known good versions for them since focus of
the daily jobs is OPNFV platform testing.
Ansible log can easily become massive depending on the level it is set
and it is not so relevant for daily jobs to have it like this due to
less probability of dailies failing to provision/deploy.
On the other hand, the jobs that verify/test bifrost and openstack-ansible
will have max verbosity enabled to troubleshoot these components themselves.
New set of jobs will be created running periodically against the HEAD
of master for bifrost and openstack-ansible to catch working versions
and help us promote them to use for daily platform jobs. Verbosity of these
jobs will be enabled like how we are doing for bifrost-verify jobs.
Change-Id: Ic9fd2ff418fdaa3ad7fbfc22d8ec2dc1174df65f
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|
|
Jobs are adjusted as well.
Change-Id: I0bc5bdda37277053d81d7da394e0156902da817a
Signed-off-by: Fatih Degirmenci <fatih.degirmenci@ericsson.com>
|