Age | Commit message (Collapse) | Author | Files | Lines |
|
Make sercurity group can be configured via the context.
The format would be:
context:
name: demo
image: yardstick-image
flavor: yardstick-flavor
user: ubuntu
security_group:
rules:
- remote_ip_prefix: "0.0.0.0/0"
protocol: "tcp"
port_range_min: 1
port_range_max: 65535
- remote_ip_prefix: "0.0.0.0/0"
protocol: "udp"
port_range_min: 1
port_range_max: 65535
- remote_ip_prefix: "0.0.0.0/0"
protocol: "icmp"
JIRA: YARDSTICK-1360
Change-Id: I00c45767ee2d70b790590e824599d5a4c274bced
Signed-off-by: rexlee8776 <limingjiang@huawei.com>
|
|
Problem
op_utils.get_volume_id takes 2 arguments and only pass 1. This would
cause tc006 fail!
volume_id = op_utils.get_volume_id(volume_name)
it will cause:
Running pylint...
You can speed this up by running it on 'HEAD~[0-9]' (e.g. HEAD~0, this
change only)...
Using config file /home/opnfv/repos/yardstick/.pylintrc
************* Module yardstick.orchestrator.heat
E:235,20: No value for argument 'volume_name' in function call
(no-value-for-parameter)
To Do
It doesn't need to call get_volume_id here, just use heat get_resource
would be enough. This fix should work with tc006 test case.
JIRA: YARDSTICK-1371
Change-Id: I913ffd085f742ae4464dbcc4269719e2c50e37b5
Signed-off-by: rexlee8776 <limingjiang@huawei.com>
|
|
|
|
JIRA: YARDSTICK-1316
Change-Id: I792c873b42aba80be62ff7a93f6d02ea89e42356
Signed-off-by: Mytnyk, Volodymyr <volodymyrx.mytnyk@intel.com>
|
|
delete_custom_resource_definition()"
|
|
|
|
JIRA: YARDSTICK-1317
Change-Id: I2dc924334199761d63dce9aebe9ce200ca0fddf8
Signed-off-by: Mytnyk, Volodymyr <volodymyrx.mytnyk@intel.com>
|
|
When creating a network the host should be checked to see if the network
is already created. If the required network is already there it should be used.
JIRA: YARDSTICK-1315
Change-Id: I09b114a728364ee56397af70cc48f1b7904f06cc
Signed-off-by: John O Loughlin <john.oloughlin@intel.com>
|
|
JIRA: YARDSTICK-1330
Change-Id: I05303a1d0783421472883806f97441dd63f7f740
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
JIRA: YARDSTICK-1321
Change-Id: Id0f3a5d93f11674fd3908a7f5cf7214afcfa3600
Signed-off-by: Mytnyk, Volodymyr <volodymyrx.mytnyk@intel.com>
|
|
Accept strings and list of strings as "args" and "commands"
in a Kubernetes container.
JIRA: YARDSTICK-1329
Change-Id: I56470741072fb7f9a62d695c51fcb0cc3f3ff1b9
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
|
|
The configuration parameter "imagePullPolicy" defines how Kubernetes
retrieves de image. By default, Kubernetes will always try to pull the
latest image from the repository [1]. Three possible arguments are accepted:
- Always
- IfNotPresent
- Never
Example of Kubernetes pod definition:
containers:
- name: vcmts-pktgen-pod-0
image: openretriever/yardstick
imagePullPolicy: IfNotPresent
Example of Yardstick definition:
context:
name: k8syardstick
type: Kubernetesservers:
vnf:
image: openretriever/yardstick
imagePullPolicy: IfNotPresent
[1] https://kubernetes.io/docs/concepts/containers/images/
JIRA: YARDSTICK-1328
Change-Id: I83bfb01670eef22218b1affd7d37af14e32a622e
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
In Kubernetes context, service NodePort can contain more than one port
defined. Actually, by default port SSH (22) is always created. If more
than one port is defined in the service template, 'name' parameter is
mandatory.
Names must be lowercase, containing alphanumeric characters or '-'.
Verification regex used by Kubernetes: [a-z0-9]([-a-z0-9]*[a-z0-9])?
JIRA: YARDSTICK-1324
Change-Id: I82791761d8eae24196c2f16aee9900af28d44c57
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
Add to "Kubernetes" context the "interfaces" information when retrieving
a server. This information is needed for NSPerf test cases.
The interface information comes from the resource controller network list.
Each replication controller will have one port per network defined.
JIRA: YARDSTICK-1303
Change-Id: Ifb0e17df295c042a643128c705a93876af999bad
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
|
|
This new parameter, "torelations", is applied to pods and allow (but not
require) the pods to schedule onto node with matching tains [1].
Example of "tolerations" definition in a Kubernetes pod:
spec:
containers: [...]
tolerations:
- key: <key defined in the taint>
value: <key value to match with the taint>
effect: <effect in case of match>
operator: <matching operator>
- key: ...
Example of "tolerations" definition in a Yardstick test case:
context:
type: Kubernetes
servers:
host:
containers: [...]
tolerations:
- key: ...
value: ...
effect: ...
operator: ...
NOTE: if any toleration is defined, a default one will be applied in
order to allow any replication controller to create the pods in any
Kubernetes node. This default toleration is defined as:
spec:
tolerations:
- operator: "Exists"
[1] https://kubernetes.io/docs/concepts/configuration/taint-and-toleration/
JIRA: YARDSTICK-1254
Change-Id: I32fb9c7086b4218c323218738057f634eb6ffff4
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
This new parameter, "ports", will allow the user to
automatically to add port variables in a pod definition:
context:
type: Kubernetes
servers:
host:
image: ...
commands: ...
ports:
- containerPort: <port value> # Mandatory
hostIP: <host IP> # Optional
hostPort: <host port> # Optional
name: <port name> # Optional
protocol: <protocol name> # Optional
The port type and the definition must be one of the supported ones
in Kubernetes [1].
Container port definition, V1 API, Kubernetes Python library v6.0.0 [2]
[1] https://kubernetes.io/docs/concepts/services-networking/connect-applications-service/
[2] https://github.com/kubernetes-client/python/blob/6.0.0/kubernetes/client/models/v1_container_port.py
JIRA: YARDSTICK-1183
Change-Id: I3673dbeec8d380cb2f8b1dacf6c25c7cdbc37e28
Signed-off-by: John O Loughlin <john.oloughlin@intel.com>
|
|
This new parameter, "restartPolicy", will allow define the restart policy per pod.
Example of yaml definition in Kubernetes:
apiVersion: v1
kind: Pod
metadata:
name: frontend
spec:
containers:
- name: db
...
restartPolicy: Always # Possible values: "Always", "OnFailure" and "Never"
Example of definition in a Yardstick context:
context:
type: Kubernetes
servers:
host:
containers:
- name: ...
restartPolicy: Always # Default value: "Always"
Restart policy [1].
[1] https://kubernetes.io/docs/concepts/workloads/pods/pod-lifecycle/#restart-policy
JIRA: YARDSTICK-1175
Change-Id: Id4317b909f98422e89d6d4553e8cfb0e1f593355
Signed-off-by: John O Loughlin <john.oloughlin@intel.com>
|
|
Like in Heat context, network elements should be defined in a
dictionary, using the network name as the entry key.
Previous implementation:
context:
networks:
- name: ...
plugin: ...
args: ...
New implementation:
context:
networks:
net_name_1:
plugin: ...
args: ...
net_name_2: ...
JIRA: YARDSTICK-1160
Change-Id: I786b06555af509440604305e2c3c2c9fe521bef8
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
This new parameter, "resources", will allow the user to automatically
add resources in a pod definition per container.
Example of yaml definition in Kubernetes:
apiVersion: v1
kind: Pod
metadata:
name: frontend
spec:
containers:
- name: db
image: mysql
resources:
requests:
memory: "64Mi"
cpu: "250m"
limits:
memory: "128Mi"
cpu: "500m"
Example of definition in a Yardstick context:
context:
type: Kubernetes
servers:
host:
resources: # There are two possible keys in this dict:
# "requests" and "limits".
limits: <dictionary>
requests: <dictionary>
Resources in containers [1].
[1] https://kubernetes.io/docs/concepts/configuration/manage-compute-resources-container/
JIRA:YARDSTICK-1171
Change-Id: I163d72d9d3e88b538fca73eb4dbf38613aa23ff4
Signed-off-by: John O Loughlin <john.oloughlin@intel.com>
|
|
This new parameter, "env", will allow the user to
automatically to add environment variables in a pod definition
Example of definition in a context:
context:
type: Kubernetes
servers:
host:
image: ...
commands: ...
env:
- <variable name>: <variable value>
- <variable name>: <variable value>
The volume type and the definition must be one of the supported
ones in Kubernetes [1].
[1] https://kubernetes.io/docs/concepts/storage/volumes/#types-of-volumes
JIRA: YARDSTICK-1161
Change-Id: Icf372a98c090063f10c9217755519ab92ba9b0ba
Signed-off-by: John O Loughlin <john.oloughlin@intel.com>
|
|
This feature will give the user the ability to create user defined "NodePort"
services per pod, a part from the default SSH port created, by default, for
each pod created.
Example of Kubernetes yaml definition:
apiVersion: v1
kind: Servicemeta
data:
name: pod-1-service-nodeport
spec:
type: NodePort
ports:
- name: web
port: 80
targetPort: 8888
nodePort: 33333
clusterIP: 10.254.0.8
selector:
app: pod-1
Example of Yardstick test case definition:
context:
type: Kubernetes
servers:
host:
containers:
- image: ...
securityContext:
allowPrivilegeEscalation: false
- image: ...
node_ports:
- port: <port number> # Mandatory
name: <port name> # Optional
targetPort: <port name> # Optional, default: targetPort=port
nodePort: <port number> # Optional, assigned by Kubernetes
Kubernetes service, type "NodePort" [1]
[1] https://kubernetes.io/docs/concepts/services-networking/service/#type-nodeport
JIRA: YARDSTICK-1181
Change-Id: Ia7730e4569d5526d573402a98c27d55c5c82cdef
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
The class name "KubernetesObject" is not accurate. This class holds
the definition of a Kubernetes replication controller. The suggested name
is therefore "ReplicationControllerObject".
JIRA: YARDSTICK-1180
Change-Id: I9ee9228f9528cb779e54a9bc178760b6e24c85c2
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
If CRD "Network" is defined and network items are created, each pod (server)
can have access to one or several networks. This is defined in the metadata
section, as "annotations.networks" [1].
Example of Kubernetes pod definition with networks:
apiVersion: v1
kind: Pod
metadata:
name: test-pod
annotations:
networks: '[{"name": "flannel"}]'
Example of Yardstick server definition with networks:
context:
type: Kubernetes
servers:
host:
containers:
- name: ...
networks:
- flannel # These names must be defined in
# context.networks
...
networks:
- name: flannel
plugin: flannel
Kubernetes annotations [2].
[1]https://github.com/intel/multus-cni/tree/b9446232cdf4f1b6f2bea583291973cc97e963f4#configuring-multus-to-use-kubeconfig-and-a-default-network
[2]https://kubernetes.io/docs/concepts/overview/working-with-objects/annotations/
JIRA: YARDSTICK-1178
Change-Id: I6e7b4bacf10810833ec733c14d44e5db613675e3
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
Add a new Kubernetes resource kind: "Network" [1] [2]
Kubernetes network plugins (alpha definition): [3]
Network definition example in Kubernetes:
apiVersion: "kubernetes.com/v1"
kind: Network
metadata:
name: flannel
plugin: flannel
args: '[
{delegate": {"isDefaultGateway": true}}
]'
Proposed Kubernetes context network definition:
context:
networks:
- name: flannel
plugin: flannel
args: (string)
- name: sriov_upload
plugin: sriov
args: (string)
[1]https://builders.intel.com/docs/networkbuilders/multiple-network-interfaces-in-kubernetes-application-note.pdf
[2]http://cdn.opensourcecloud.cn/zt/k8s/01.pdf
[3]https://kubernetes.io/docs/concepts/cluster-administration/network-plugins/
JIRA: YARDSTICK-1160
Change-Id: I71a49ac14e3d28ded91d2ed3cd9cc527e40303f7
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
Custom resource definition example in Kubernetes:
apiVersion: "apiextensions.k8s.io/v1beta"
kind: CustomResourceDefinition
metadata:
name: networks.kubernetes.com
spec:
group: kubernetes.com
version: v1
scope: Namespaced
names:
plural: networks
singular: network
kind: Network
Proposed Kubernetes context network definition:
context:
custom_resources:
- name: network # name of the resource (singular)
version: v1 # optional, "v1" by default
scope: Namespaced # optional, "Namespaced" by default
From this definition, we will extract the Kubernetes parameters:
- metadata.name: custom_resources.name + "s" + context_name + ".com"
- spec.group: context_name + ".com"
- spec.scope: custom_resources.scope
- spec.version: custom_resources.version
- spec.names.plural: custom_resources.name + "s"
- spec.names.singular: custom_resources.name
- spec.names.kind: custom_resources.name with first capital letter
[1] https://kubernetes.io/docs/concepts/api-extension/custom-resources/
JIRA: YARDSTICK-1163
Change-Id: If8980dc3f6ddf9c6949bf15be8011aa98482ddc9
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
This new parameter, "securityContext", will allow the user to define the
privilege and access control settings for a pod or a container [1].
Example of "securityContext" definition in the pod (if only one container
is defined):
context:
type: Kubernetes
servers:
host:
image: ...
securityContext:
runAsUser: 1000
fsGroup: 2000
Example of "securityContext" definition in the pod and the container (if
several containers are defined):
context:
type: Kubernetes
servers:
host:
securityContext:
runAsUser: 1000
fsGroup: 2000
containers:
- image: ...
securityContext:
allowPrivilegeEscalation: false
- image: ...
[1] https://kubernetes.io/docs/tasks/configure-pod-container/security-context/
JIRA: YARDSTICK-1156
Change-Id: I597a300c68cd834522a284b1cca0faa918493342
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
Add the ability to define not only one but many containers per pod in a
replication controller descriptor. This feature must be backwards
compatible; all current test cases using the "single container" server
definition must be accepted.
Example of single container pod definition:
context:
type: Kubernetes
servers:
host:
image: ...
commands: ...
volumes:
- name: volume1 # mandatory
<volume type definition> # mandatory
Example of several container pod definition:
context:
type: Kubernetes
servers:
host:
containers: # if this key is present, all container specific
parameters (image, commands, args, volumeMounts, etc.)
must be defined per container
- image: ...
commands: ...
- image: ...
commands: ...
volumes:
- name: volume1 # mandatory
<volume type definition> # mandatory
NOTE: other parameters, like "volumes" or "nodeSelector", are common to all
containers in the pod.
JIRA: YARDSTICK-1155
Change-Id: Ib95668c68e9c09e6de3f1aa41c903cc52e6809ad
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
This new parameter, "volumeMounts", will allow the user to automatically
create new volumes mounted inside the container. Example of Kubernetes
context definition:
context:
type: Kubernetes
servers:
host:
volumeMounts:
- name: volume1 # mandatory
mountPath: /dev/hugepages # mandatory
readOnly: True # optional, default=False
JIRA: YARDSTICK-1151
Change-Id: Ic00b45e6d603a9f85e3f0e25becdf0bce14e637c
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
This new parameter, "volumes", will allow the user to automatically create
new volumes. Example of Kubernetes context definition:
context:
type: Kubernetes
servers:
host:
image: ...
commands: ...
volumes:
- name: volume1 # mandatory
<volume type definition> # mandatory
The volume type and the definition must be one of the supported ones in
Kubernetes [1].
[1] https://kubernetes.io/docs/concepts/storage/volumes/#types-of-volumes
JIRA: YARDSTICK-1152
Change-Id: I44a91c605f047de4f286407e28fb5aa2e921b00a
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
To access to the container without using a password, the jumphost
RSA public key is copied to each container, using "volumeMounts"
defined as "configMap", to /root/.ssh/authorized_keys.
To work properly, the following permissions must be set:
- /root/.ssh: 700
- /root/.ssh/authorized_keys: 600
Because of [1][2], the mounted folders have fixed permissions and
cannot be modified.
[1]https://groups.google.com/forum/#!topic/kubernetes-dev/eTnfMJSqmaM
[2]https://github.com/kubernetes/kubernetes/issues/28317
JIRA: YARDSTICK-1149
Change-Id: I821064da56699c5b4f509d233c33e55af119fd56
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
This new parameter will contain the OpenStack cloud specific configuration
used by Shade client. This new flag is used only in Heat context.
By default, this new parameter (dict) will contain this content:
'os_cloud_config': {'verify': False}
This field will be used by HeatStack [1] to create a Shade cloud. Shade
retrieves, if not defined, the OpenStack configuration from
"os_client_config". This configuration is used to generate the cloud
configuration, which is the description of the OpenStackCloud returned.
The default parameter defined, "verify", refers to the related bug. By
default, in case of using SSL certificate it will not be verified.
[1] https://github.com/opnfv/yardstick/blob/b338d3091bb0beb89d4ad9f7c144f43a31a19a74/yardstick/orchestrator/heat.py#L47
JIRA: YARDSTICK-1139
Change-Id: I875a7018401b84e51dab775b8194174645d27e06
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
We need to see what Heat failures occured
so that we can debug CI failures.
The only way to do this seems to be to use the
private event_utils.get_events function to
query for all failures
Also since we had a failure go ahead and dump the
Heat template that failed.
Sample output:
2018-03-25 23:50:08,765 [INFO] yardstick.orchestrator.heat heat.py:629 Creating stack 'yardstick-460ed969' START
2018-03-25 23:50:21,932 [ERROR] yardstick.orchestrator.heat heat.py:644 Resource CREATE failed: BadRequest: resources.yardstick-460ed969-xe1: Invalid input for operation: physical_network 'nosuch' unknown for flat provider network.
Neutron server returns request_ids: ['req-6f981f1e-a9e2-4114-af84-1ee528aed51b']
2018-03-25 23:50:21,933 [ERROR] yardstick.orchestrator.heat heat.py:644 BadRequest: resources.yardstick-460ed969-xe1: Invalid input for operation: physical_network 'nosuch' unknown for flat provider network.
Neutron server returns request_ids: ['req-6f981f1e-a9e2-4114-af84-1ee528aed51b']
2018-03-25 23:50:21,972 [ERROR] yardstick.orchestrator.heat heat.py:645 {'description': '\n'
'All referred generated resources are prefixed with the '
'template\n'
'name (i.e. yardstick-460ed969).\n',
'heat_template_version': '2013-05-23',
'outputs': {'trafficgen_1.yardstick-460ed969': {'description': 'VM UUID',
'value': {'get_resource': 'trafficgen_1.yardstick-460ed969'}},
'trafficgen_1.yardstick-460ed969-fip': {'description': 'floating '
'ip '
JIRA: YARDSTICK-998
Change-Id: Ia8f4e5ba7e280fb9086519680d5ee90a2b442e6b
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
|
|
|
|
|
|
|
|
* Add yardstick/orchestrator/heat.py:HeatStack.get
* Add yardstick/benchmark/contexts/heay.py:HeatContext.retrieve_existing_stack
JIRA: YARDSTICK-886
Change-Id: I6974b79a25f98066a49b1bc8ccd11383e7962091
Signed-off-by: Emma Foley <emma.l.foley@intel.com>
|
|
context.key_uuid is only used as a suffix for the key files.
There is already a unique ID associated with a the context;
the context name includes a task ID, which is a UUID.
This patch sets context.key_uuid to to context.name instead of
generating a separate UUID
As a result, get_short_key_uuid() is not needed.
JIRA: YARDSTICK-1028
Change-Id: Id175061d6cfe23a068bb3d12ce176c1f176e8236
Signed-off-by: Emma Foley <emma.l.foley@intel.com>
|
|
JIRA: YARDSTICK-1034
Change-Id: I613b8bff669fceeb0c7c59dec2b6f1ab22b09247
Signed-off-by: mting <dianfeng.du@ericsson.com>
|
|
JIRA: YARDSTICK-970
Change-Id: I593af4b3a30e52521e3d77fd84bec16ab4733483
Signed-off-by: rexlee8776 <limingjiang@huawei.com>
|
|
There is a bug in Shade, in the function "event_utils.poll_for_events".
While Shade is waiting for the Heat stack to be deleted, retrieves
periodically the status of this Heat stack. When the stack is deleted,
the function "get_stack" [1] returns None; then L90 tries to access to
a key of this None object.
[1]https://github.com/openstack-infra/shade/blob/e5bcb705c5b32334cdc1dd9807126298c87b0622/shade/_heat/event_utils.py#L89-L90
Shade function
JIRA: YARDSTICK-1015
Change-Id: I56ea853be41bfd7f63e094d7a3696d97ba078ba4
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
we need to use jsonutils.dump_as_bytes for NamedTemporaryFile.write()
2018-02-11 19:15:01,366 [INFO] yardstick.orchestrator.heat heat.py:563 Creating stack 'yardstick-4359f33e' START
2018-02-11 19:15:01,935 [ERROR] yardstick.benchmark.contexts.heat heat.py:317 stack failed
Traceback (most recent call last):
File "/home/rbbratta/yardstick-upstream/yardstick/yardstick/benchmark/contexts/heat.py", line 313, in deploy
timeout=self.heat_timeout)
File "/home/rbbratta/yardstick-upstream/yardstick/yardstick/orchestrator/heat.py", line 567, in create
stack.create(self._template, self.heat_parameters, block, timeout)
File "/home/rbbratta/yardstick-upstream/yardstick/yardstick/orchestrator/heat.py", line 56, in create
template_file.write(jsonutils.dumps(template))
File "/home/rbbratta/yardstick_venv3/lib/python3.5/tempfile.py", line 622, in func_wrapper
return func(*args, **kwargs)
TypeError: a bytes-like object is required, not 'str'
JIRA: YARDSTICK-1005
Change-Id: I504f23b86119b62f3aea5b83a445b97bf810220c
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
|
|
Replaced "python-heatclient" with "shade" client. The "python-heatclient"
library is removed from requirements file. "shade" client version 1.22.2
is the required one for OpenStack Pike release.
shade is the recommended client to interact with OpenStack clouds [1]
[1] https://github.com/openstack-infra/shade
JIRA: YARDSTICK-880
Change-Id: Ibbdfc42a543d2d9ed4cf70d78de3de8cbecd0076
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>
|
|
|
|
When creating a Neutron security group for a heat template. Add
IPv6 security group rules to access the resouces.
Change-Id: I0b541dfe400121e35908a70778665ee1576f0622
Signed-off-by: Daniel Martin Buckley <daniel.m.buckley@intel.com>
|
|
Users would like to ensure placement of VMs on specific compute nodes so
that the measurements are meaningful. Examples: Measure network
performance in different scenarios (VMs in same host, in different
hosts, across fabric, across tenants)
Example:
context:
name: yardstick
placement_groups:
pgrp1:
policy: "availability"
servers:
tg_0:
floating_ip: true
placement: "pgrp1"
availability_zone: "zone2"
vnf_0:
floating_ip: true
placement: "pgrp1"
availability_zone: "zone1"
Change-Id: I28a757c25ae3f5b3571ab3edd82d51ceba32c302
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
|
|
node when creating containers for Kubernetes context"
|
|
JIRA: YARDSTICK-803
Currently kubernetes test case can only run in master node.
We need to support it run in jump server.
So I add service and use nodePort type.
Then we can login the pod using nodePort.
Change-Id: Ia7900d263f1c5323f132435addec27ad10547ef9
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>
|
|
Kubernetes running node when creating containers for
Kubernetes context
For example, a yaml file may looks like:
servers:
host:
image: xxx
command: /bin/bash
nodeSelector:
xxx: yyy
Synchronously change the unit test for this function
Change-Id: If74c9dad9b1a70395bb79f34708a0fde04e7e650
Signed-off-by: Trevor Tao <trevor.tao@arm.com>
|