Extending the Neutron with configuration parameters and scripts for
deploying OVS with DPDK support enabled. The new functionality takes
care of binding NICs to DPDK and adding those to OVS bridges of type
'netdev'.
Co-Authored-By: Rihab Banday <rihab.banday@ericsson.com>
Change-Id: I9932123986a0b723d7523136940d325bcfde983d
This adds the rabbitmq statefulset name override to the values
for each openstack service chart in the Armada manifest, as this
is required to support communicating with the rabbitmq backends
directly
Change-Id: I4a2fabf491c6607bec80f37a8d1236c55ec8430a
Signed-off-by: Steve Wilkerson <sw5822@att.com>
This disables the cephfs provisioner in the multinode
periodic jobs. It seems the helm tests for the ceph
provisioner chart that test cephfs fail more often than
not in the multinode jobs while passing reliably in the
single node check and gate jobs. As cephfs is still
gated, disabling the cephfs provisioner in the periodic
jobs allows for further investigation into this issue
without causing potential regressions
Change-Id: Id4cf7d6571855413be4b67a440423445af3270e3
Signed-off-by: Steve Wilkerson <sw5822@att.com>
Currently OSH_EXTRA_HELM_ARGS_CEPH is calculated for ceph-mon only and
the same value is used for other charts from loop.
Change-Id: Ia014310ab675e4f58c224aae1139dfb3a065766b
In the 900-use-it.sh script, the user can define the gateway
(OSH_BR_EX_ADDR) and the subnet of the external network
(OSH_EXT_SUBNET). However, in the script that configures the gateway
correctly in br-ex, these values are hardcoded.
This patch allows the user to modify them by a env. variable, same as
the 900-use-it.sh script does
Change-Id: I25124024ebb72ff5fea8903fb393b71a052d0f93
Signed-off-by: Manuel Buil <mbuil@suse.com>
This ps updates the component check scripts to deploy
the default number of rabbit replicas in the gate
Change-Id: I996af138373476ae4a65dc2bef7e360155390a5d
Wrap code making the assumption there is only one Ceph backend
(using is_ceph_volume_configured) in a "range" and use
ceph_backend_list helper to iterate all available Ceph backends.
Move Ceph pool application name in values.yaml from
conf.software.rbd.rbd_pool_app_name* to conf.ceph.pools.*.app_name
Change-Id: If1126e51fe9ebb85185e375dc282e83db63d934c
Depends-On: Iaa67061b05a9d355228ad7d3f5ee0f4f04dbdc66
Signed-off-by: Daniel Badea <daniel.badea@windriver.com>
The name of the variable ends with LIBVIRT although it should end with
OPENVSWITCH, as used in line 29
Change-Id: I0ec490de44969aea2cccf51d753f1ffcfa9e2511
Signed-off-by: Manuel Buil <mbuil@suse.com>
Currently, in the script which sets up VMs with vmbc instead of BM for ironic
practice,
There is an IF command checking if OVS_VSWITCHD pod is located in master
node or not.
To get the info about Pod's nodeName, command below has been used.
$ kubectl --namespace openstack get pod "${OVS_VSWITCHD_POD}" -o wide --no-headers | awk '{ print $NF }'
But this command sometimes cannot parse Pod's nodeName correctly becuase
new headers such as "NOMINATED NODE" and "READINESS GATE" are printed in the end "-o wide" option.
(in some version of k8s)
To avoid this problem and give more readability,
I suggest changing commands to use "-o json" option.
Change-Id: Ib81ae505ac8d04dae9af6326880e1fa17664ac0b
On some deployments the tests run on Neutron for multinode deployments
exceed the current 15 minutes timeout. The timeout is now configurable
through the OSH_TEST_TIMEOUT variable which can be passed from the yaml
file. The default timeout remains 15 minutes.
Change-Id: Ic5a1e9fd812e51ed51aef4de162b1c08d1c9a593
Signed-off-by: Abel Navarro <abel.navarro@gmail.com>
- As part of PS 634782, Ceph-rgw chart was updated. We need to update
Ceph chart install script as per new changes.
Change-Id: Ieede912272903ed251693879f17a821cb2a4e7bb
This PS adds checks for the Stein Release of OpenStack in Ubuntu Bionic
containers.
Depends-On: https://review.opendev.org/667726
Change-Id: Icfad3434ca496a841993b95adaf5d853728d920f
Signed-off-by: Pete Birley <pete@port.direct>
This PS indroduces a simpler way to incorp over-rides into gate
runs, and also ensures that they are scoped to a single chart, rather
than all of the charts deployed within a gate run.
Depends-On: https://review.opendev.org/666957
Change-Id: I49edf52cc1fc5ec60ee9754c28880c9c0c54492e
Signed-off-by: Pete Birley <pete@port.direct>
This PS updates the charts to use the htk function recently introduced
to allow oslo.messaging clients ans servers to directly hit their
backends rather than using either DNS or K8S svc based routing.
Depends-On: I5150a64bd29fa062e30496c1f2127de138322863
Change-Id: I458b4313c57fc50c8181cedeca9919670487926a
Signed-off-by: Pete Birley <pete@port.direct>
This PS updates the deplyment script for the compute-kit to allow use
from outside the k8s cluster.
Change-Id: Ie758b3c3c6b36c1882030c9a1be1b2ba6cc13ed4
Signed-off-by: Pete Birley <pete@port.direct>
This adds a helm test for Horizon, the helm test
runs a selenium webdriver check to verify the dashboard
is up
Change-Id: I3616c05596b2bd94931c39fb774333bf65453d52
Signed-off-by: Steve Wilkerson <sw5822@att.com>
This change removes the network policy overrides from the default
developer deployment scripts now that there is a set of
designated scripts that provide this functionality.
Change-Id: I788845e85314b3a8a9c860fbc77de536113781e0
This removes the logging overrides for using the oslo log
fluentd handler and formatter from the armada-osh manifest. As
the armada jobs in the openstack-helm repo no longer deploy the
lma components, there's no fluentd instance for the openstack
services to log to
Change-Id: I6aed11871046a2caf705a3c444d169d1d977bdee
Signed-off-by: Steve Wilkerson <sw5822@att.com>
The volume variables is changed in mariadb chart.
It affects openstack helm gating job of mariadb, so we need to
change varible to
volume.use_local_path_for_single_pod_cluster.enabled.
Enable the local host path for mariadb data so if the container
is restarted during the gate jobs it can recover successfully.
Story: 2004702
Task: 33462
Depends-on: https://review.opendev.org/661955
Change-Id: Ib55779ddcbe7033da1900f5cd21de46543392d11
Prometheus mysql exporter is failing because of missing label in
network policy for mysql. This is to add the label to the override
yaml
Change-Id: Ic5e2613303f04d2bfab1229d3818a2fc9b4263bc
Signed-off-by: Sreejith Punnapuzha <Sreejith.Punnapuzha@outlook.com>
This adds missing passwords to the armada-osh manifest. These
passwords include: passwords for the test users used for the
openstack chart helm tests, the password for the mariadb sst user
deployed in the openstack namespace, and the mariadb exporter
user for the openstack mariadb instance
Change-Id: Ia04ab3d83d1967c4ab65915adaec53d488f451fb
Create the overrides file necessary for Rocky release:
- api-paste entrypoint neutron.api.versions:Versions.factory deprecated
Deployment script needs to be executed with OSH_OPENSTACK_RELEASE value.
Change-Id: If0d3553bd004426d8e97e1fa62ee9b99f4a895a9
