MessagingConsumer accepts messages from multiple producers

The messaging consumer now can store a list of PID of several producers.
The notification handler can compare the procedence of a message from
a list of PID.

JIRA: YARDSTICK-1074

Change-Id: I193f83c2b471e5bf1298ac728be52533aded0c1a
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Merge "add verbose to debug losetup delete"

Merge "remove apex lake tc007 from test suites"

Merge "Replace neutron create security group full with shade."

remove apex lake tc007 from test suites

fixes opnfv_os-nosdn-fdio-noha_daily.yaml failures due
to missing testcase

Change-Id: I5f050d72334da46b94de5fa3662047bf694f7e3e
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

fix unittest test_deploy_try_retrieve_context_does_not_exist

unittest is failing due to mock being called multiple times.

I assume this is due to parallelization of tests in Jenkins somehow

======================================================================
FAIL: test_deploy_try_retrieve_context_does_not_exist (benchmark.contexts.test_heat.HeatContextTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/jenkins/opnfv/slave_root/workspace/yardstick-verify-master/.tox/py3/lib/python3.5/site-packages/mock/mock.py", line 1305, in patched
    return func(*args, **keywargs)
  File "/home/jenkins/opnfv/slave_root/workspace/yardstick-verify-master/yardstick/tests/unit/benchmark/contexts/test_heat.py", line 309, in test_deploy_try_retrieve_context_does_not_exist
    mock_path_exists.assert_called_once_with(key_filename)
  File "/home/jenkins/opnfv/slave_root/workspace/yardstick-verify-master/.tox/py3/lib/python3.5/site-packages/mock/mock.py", line 947, in assert_called_once_with
    raise AssertionError(msg)
AssertionError: Expected 'exists' to be called once. Called 3 times.

----------------------------------------------------------------------

Change-Id: I696c63ef9f7943b841e2d3d533c92f18aa7044da
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Replace neutron create security group full with shade.

Function create_security_group_full now uses shade client.

Remove openstack util's private 'get_security_groups',
'get_security_group_id' & 'create_security_group' functions as
'create_security_group_full' now uses 'get_security_group' &
'create_security_group' function defined in shade client library.

JIRA: YARDSTICK-890

Change-Id: Ibd5b9ca9b7ce99b3e40c5dcc545f2af672ce2203
Signed-off-by: Shobhi Jain <shobhi.jain@intel.com>

Merge "Remove tool provisioning in Standalone contexts"

Merge "Bugfix: interface down HA tc050 attack and monitor logistic wrong"

Add MQ consumer, producer and payload base classes

Added MessagingProducer base class. A class implementing this base class
can send a cast message using the MQ service installed in the Yardstick
jumphost (RabbitMQ by default, other MQ could be implemented). The producer
will send messages to an specific topic.

Added MessagingConsumer base class. A class implementing this base class
will be able to receive any message sent by a MessagingPorducer class
publishing messages in the topic subscribed.

By default both Producer and Consumer "fanout" is True. That means every
Consumer will create a fanout Queue attached to the MQ Exchange topic. All
Consumers attached to this topic will receive the message sent by the
Producer [1].

Added Payload base class. To send data through the message queue service,
a Payload derived object should be created. This base class allows to
define the attributes container in the class, convert the object to a dict
and retrieve the object from a dict.

Added a new library, "oslo.messaging", to implement the RPC client and
server.

[1]http://blog.thedigitalcatonline.com/blog/2013/08/21/some-tips-about-amqp-direct-exchanges/

JIRA: YARDSTICK-1074

Change-Id: I63932b5fb3de2bdc1270fc83295630a2a349e2a6
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Bugfix: interface down HA tc050 attack and monitor logistic wrong

1. simplify the attack to be one interface
2. move monitor step before attack
3. update rst doc to match

JIRA: YARDSTICK-1120

Change-Id: Icfa21d58e7ccdbeaca8a9a1a3098cceb21c981c3
Signed-off-by: rexlee8776 <limingjiang@huawei.com>

Merge "Bugfix: to list all used loop device"

Merge "Bugfix: Can't get image list in API"

Bugfix: to list all used loop device

Bug fix in script "build_yardstick_image.yml", to

- Use option "-a" to list all used loop devices, to be more compatible.
- Ignore the error and continue, if the debug command failed.

JIRA: YARDSTICK-1121

Change-Id: I24aaebb2ffe3107ca5c619401725aa8c8e1af295
Signed-off-by: ting wu <ting.wu@enea.com>

Bugfix: Can't get image list in API

JIRA: YARDSTICK-1110

To match OpenStack Pike release, we upgrade novaclient from 7.1.1 to 9.1.1.
And the client interface has changed, it no longer has nova_client.images attribute.
So I use shade instead. Later will change all same issue.

Change-Id: Ie4f54069d4346e44e2ad925439930504b945cbad
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>

Merge "Bug fix in using Unixbench"

Upgrade collectd to version 5.8

Update collectd version to 5.8 as it's the latest stable
and supports GCC 7

JIRA: YARDSTICK-1114

Change-Id: I9147b7277d87047eda729664d5ee3e8ab2c145af
Signed-off-by: John O Loughlin <john.oloughlin@intel.com>

Merge "Add support for dpdk 17.05"

Add support for dpdk 17.05

Add support for dpdk 17.05. Add option to dpdk scripts
to download dpdk 17.08 17.11 and 18.02. These are currently
unsupported due to prox build issues.

JIRA: YARDSTICK-1113

Change-Id: I585566364a4d1f04a696439af96c50c0e9d38033
Signed-off-by: John O Loughlin <john.oloughlin@intel.com>

Bug fix in using Unixbench

JIRA: YARDSTICK-1117

This bug will cause the error in running unixbench.
"no such file named ./run"
So we have to fix it.

Change-Id: I141fcb5567fbb0e3df97aee677cb946a25a9788e
Signed-off-by: liyin <liyin11@huawei.com>

Bugfix: k8 scenario fix

JIRA: YARDSTICK-1109

Change-Id: Ic61a63099c3d2d4af5fa5634299e4b2946d53eaf
Signed-off-by: rexlee8776 <limingjiang@huawei.com>
(cherry picked from commit 3b1316647b50bc3bbb1eb6216aa07fd4dff34edf)

docker/supervisor.sh: Fix condition block end

Fixes: 18546b4f

Change-Id: I51fcb48e790ff7c57273a3f6712ccb2b9bc4aa32
Signed-off-by: Alexandru Avadanii <Alexandru.Avadanii@enea.com>
(cherry picked from commit ead90064d12be552bd334828d41ffffa75fad01b)

docker/supervisor.sh: Fix condition block end

Fixes: 18546b4f

Change-Id: I51fcb48e790ff7c57273a3f6712ccb2b9bc4aa32
Signed-off-by: Alexandru Avadanii <Alexandru.Avadanii@enea.com>

[AArch64] Dockerfile: Fix malformed patch

Change-Id: Ia2dfc6c4def8e71e36747b8f27cbf22700f676b1
Signed-off-by: Alexandru Avadanii <Alexandru.Avadanii@enea.com>
(cherry picked from commit 161f7d676b4404a5445681a7871e086b751fdafc)

[AArch64] Dockerfile: Fix malformed patch

Change-Id: Ia2dfc6c4def8e71e36747b8f27cbf22700f676b1
Signed-off-by: Alexandru Avadanii <Alexandru.Avadanii@enea.com>

Merge "Bugfix: k8 scenario fix"

[userguide] Use literalinclude directive in 05-operation

Uses literalinclude directive instead of having a copy of the file content in documentation.
This means that any update to this file is detected, and the most up-to-date content is used.

If a testcase is updated, the updates wil be picked up.
The examples used are ones that are tested reqularly.

JIRA: YARDSTICK-1066

Change-Id: I15d967eb04d1e6f4348683d8746fa9989dceda94
Signed-off-by: Emma Foley <emma.l.foley@intel.com>

[docs] update the user guide for "yardstick-operation"

* Add a new chapter: operation which shows how to use Yardstick and
  how to create test cases
* Include details on how to reuse and keep VNFs between testcases

JIRA: YARDSTICK-885
JIRA: YARDSTICK-886
Change-Id: I365152e8812427a3c54259adbc18d5ddb216d775
Signed-off-by: Emma Foley <emma.l.foley@intel.com>

[docs] Move user guide chapters to make room for new chapter.

* Increment the numbering on chapters
* Fix some of the changes flagged by doc8

Change-Id: Ie03d165297671dfd50b9de339612cdc8bf00b988
Signed-off-by: Emma Foley <emma.l.foley@intel.com>

Bugfix: k8 scenario fix

JIRA: YARDSTICK-1109

Change-Id: Ic61a63099c3d2d4af5fa5634299e4b2946d53eaf
Signed-off-by: rexlee8776 <limingjiang@huawei.com>

ADD tc042 into CI

JIRA: YARDSTICK-1108

We add this test case in ovs_dpdk-ha and ovs_dpdk-noha testsuites
also add image build into load_images script

Change-Id: I2b0c6b106dd98c3693df18dba46259ff8ef0a76e
Signed-off-by: liyin <liyin11@huawei.com>

add verbose to debug losetup delete

JIRA: YARDSTICK-1096
JIRA: YARDSTICK-1054

Change-Id: I5a11dbd2638782c5e499cb94dd5882b8bf696e8d
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Merge "Fix RabbitMQ service installation and initialization"

Merge "don't print exceptions in atexit handler"

Merge "task: don't hide exceptions in Task"

don't print exceptions in atexit handler

atexit handler calls terminate_all after the regular python execution
path, and it looks like the traceback stack is None somehow.

In this context log.debug("", exc_info=True) doesn't work
it prints out NoneType on Python3 and
causes other problems on Python2.7

remove the exc_info=True from the logging call

JIRA: YARDSTICK-1107

Change-Id: Ida0a0ced7ff5e017e2f8608880e3bb531724af95
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

add new Fraser scenarios

New scenarios from wiki: https://wiki.opnfv.org/display/SWREL/Fraser+Scenario+Statu

Change-Id: Ifd6e45e73be2bbb99743aa3f4981d22899aab92a
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Merge "restore Heat stack failure logs for CI"

Remove tool provisioning in Standalone contexts

In both Standalone contexts, SR-IOV and DPDK, the tool provisioning must
be removed. If the tool, "dpdk-devbind.py", is not in the VNF, the test
should fail because this tool should be provided during the provisioning,
not during the test execution.

JIRA: YARDSTICK-1092

Change-Id: I212c1303fd655f151e9c0efc29aad97fe6d71d1c
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Fix RabbitMQ service installation and initialization

RabbitMQ service added in [1] is not correctly installed and initialized:
- There is an error during the installation process ("\" character
  missing).
- In the installation script, the service needs to be started first.
- In the container installation, the service needs to be started via
  supervisor.

[1]https://gerrit.opnfv.org/gerrit/#/c/53597/

JIRA: YARDSTICK-1103

Change-Id: Iade3d6ce4b522e6f576af71b7afe5559081f7929
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Merge "Replace neutron create security group rule with shade."

Merge "Dashboard for TC056"

task: don't hide exceptions in Task

If the Task raised an exception we currently hide it
and replace it with RuntimeError.  This is bad.

If an exception occured, then we don't have a result so
re-raise the original exception.

Or we could log the traceback and raise RuntimeError, but
that doesn't seem to be a good idea.

Sample traceback after re-raising original.  Without this patch the ValueError is only written
to _write_error_data

2018-03-25 22:57:56,511 yardstick.benchmark.contexts.node node.py:85 DEBUG BareMetals: []
2018-03-25 22:57:56,511 yardstick.benchmark.contexts.node node.py:89 DEBUG Env: {}
2018-03-25 22:57:56,511 yardstick.cmd.commands.task task.py:57 INFO Task FAILED
Traceback (most recent call last):
  File "/home/rbbratta/yardstick-upstream/yardstick/yardstick/cmd/commands/task.py", line 54, in do_start
    result = Task().start(param, **kwargs)
  File "/home/rbbratta/yardstick-upstream/yardstick/yardstick/benchmark/core/task.py", line 103, in start
    task_args_fnames)
  File "/home/rbbratta/yardstick-upstream/yardstick/yardstick/benchmark/core/task.py", line 321, in _parse_tasks
    task_args_fnames[i]
  File "/home/rbbratta/yardstick-upstream/yardstick/yardstick/benchmark/core/task.py", line 558, in parse_task
    context.init(cfg_attrs)
  File "/home/rbbratta/yardstick-upstream/yardstick/yardstick/benchmark/contexts/heat.py", line 131, in init
    server = Server(name, self, server_attrs)
  File "/home/rbbratta/yardstick-upstream/yardstick/yardstick/benchmark/contexts/model.py", line 210, in __init__
    (name, p))
ValueError: server 'trafficgen_1', placement 'pgrp2' is invalid
2018-03-25 22:57:56,512 yardstick.cmd.commands.task task.py:62 INFO Task FAILED
2018-03-25 22:57:56,662 yardstick.benchmark.runners.base base.py:124 DEBUG Terminating all runners
NoneType

JIRA: YARDSTICK-1102

Change-Id: I7e6fa41fc1d36f6d438a1602ab60cb41ffbee1e9
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

restore Heat stack failure logs for CI

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>

Merge "Improve "Libvirt.virsh_destroy_vm" function"

Merge "Improve "Libvirt.virsh_create_vm" function"

Merge "Improve NSB Standalone XML generation"

Merge "Improve OVS-DPDK setup process"

Merge "Improve OVS-DPDK boot process"

Replace neutron create security group rule with shade.

Rename create_secgroup_rule with create_security_group_rule.
Function create_security_group_rule now uses shade client.

JIRA: YARDSTICK-890

Change-Id: Ie0ebac67a281e55dc95c0e3e33ba43de80aba9ec
Signed-off-by: Shobhi Jain <shobhi.jain@intel.com>

Merge "Replace neutron floating ip deletion with shade."

Merge "Replace neutron floating ip creation with shade."

Merge "Replace neutron router interface deletion with shade."

Replace neutron floating ip deletion with shade.

Function delete_floating_ip now uses shade client.

JIRA: YARDSTICK-890

Change-Id: I960630926b664266afbe7be00bb1352243b41be0
Signed-off-by: Shobhi Jain <shobhi.jain@intel.com>

Merge "env: validate installer_ip from environment"

Merge "Install RabitMQ for RPC messaging between processes"

Merge "Test case spec for SDN controller resilience in non-HA config"

Merge "ansible: disable Extra cloud image kernel stub"

Merge "use basename to delete yardstick image loop devices"

Merge "TC042 bug fix"

Merge "Test case description and configuration file for yardstick_tc090:Control Node Openstack Service High Availability - Database Instances"

use basename to delete yardstick image loop devices

the full path probably isn't matching, so just grep
for the basename

JIRA: YARDSTICK-1096
JIRA: YARDSTICK-1054

Change-Id: I403a7f51310c0856fae0f79d115ba0786b7c417c
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Merge "Added scale-up documentation"

Merge "Bugfix: 'module' object has no attribute 'UnsupportedExtension'"

Merge "Add "Scenario" class wait methods for runners"

Merge "Make "Scenario" class abstract"

Bugfix: 'module' object has no attribute 'UnsupportedExtension'

JIRA: YARDSTICK-1100

This Error occur when run kuberntes test case:
The Key is the pyopenssl dependency, it seems
YARDSTICK-1032(https://jira.opnfv.org/browse/YARDSTICK-1032) encounter
the same problem. Downgrade requests from 2.18.2 to 2.11.1 would solve
this problem.

Here it is the error log:
Traceback (most recent call last):
    File "/usr/lib/python2.7/atexit.py", line 24, in _run_exitfuncs
        func(*targs, **kargs)
        File
    "/home/opnfv/repos/yardstick/yardstick/benchmark/core/task.py",
        line 301, in atexit_handler
        context.undeploy()
        File
        "/home/opnfv/repos/yardstick/yardstick/benchmark/contexts/kubernetes.py",
        line 63, in undeploy
        self._delete_ssh_key()
        File
        "/home/opnfv/repos/yardstick/yardstick/benchmark/contexts/kubernetes.py",
        line 133, in _delete_ssh_key
        k8s_utils.delete_config_map(self.ssh_key)
        File
        "/home/opnfv/repos/yardstick/yardstick/common/kubernetes_utils.py",
        line 179, in delete_config_map
        **kwargs)
        File
        "/usr/local/lib/python2.7/dist-packages/kubernetes/client/apis/core_v1_api.py",
        line 9059, in delete_namespaced_config_map
        (data) =
    self.delete_namespaced_config_map_with_http_info(name,
                        namespace, body, **kwargs)
        File
        "/usr/local/lib/python2.7/dist-packages/kubernetes/client/apis/core_v1_api.py",
        line 9159, in delete_namespaced_config_map_with_http_info
        collection_formats=collection_formats)
        File
        "/usr/local/lib/python2.7/dist-packages/kubernetes/client/api_client.py",
        line 321, in call_api
        _return_http_data_only, collection_formats,
        _preload_content,
            _request_timeout)
        File
        "/usr/local/lib/python2.7/dist-packages/kubernetes/client/api_client.py",
        line 155, in __call_api
        _request_timeout=_request_timeout)
        File
        "/usr/local/lib/python2.7/dist-packages/kubernetes/client/api_client.py",
        line 387, in request
        body=body)
        File
        "/usr/local/lib/python2.7/dist-packages/kubernetes/client/rest.py",
        line 256, in DELETE
        body=body)
        File
        "/usr/local/lib/python2.7/dist-packages/kubernetes/client/rest.py",
        line 166, in request
        headers=headers)
        File
    "/usr/local/lib/python2.7/dist-packages/urllib3/request.py",
        line 66, in request
        **urlopen_kw)
        File
    "/usr/local/lib/python2.7/dist-packages/urllib3/request.py",
        line 87, in request_encode_url
        return self.urlopen(method, url, **extra_kw)
        File
        "/usr/local/lib/python2.7/dist-packages/urllib3/poolmanager.py",
        line 321, in urlopen
        response = conn.urlopen(method, u.request_uri, **kw)
        File
        "/usr/local/lib/python2.7/dist-packages/urllib3/connectionpool.py",
        line 601, in urlopen
        chunked=chunked)
        File
        "/usr/local/lib/python2.7/dist-packages/urllib3/connectionpool.py",
        line 346, in _make_request
        self._validate_conn(conn)
        File
        "/usr/local/lib/python2.7/dist-packages/urllib3/connectionpool.py",
        line 850, in _validate_conn
        conn.connect()
        File
    "/usr/local/lib/python2.7/dist-packages/urllib3/connection.py",
        line 337, in connect
        cert = self.sock.getpeercert()
        File
        "/usr/local/lib/python2.7/dist-packages/urllib3/contrib/pyopenssl.py",
        line 348, in getpeercert
        'subjectAltName': get_subj_alt_name(x509)
        File
        "/usr/local/lib/python2.7/dist-packages/urllib3/contrib/pyopenssl.py",
        line 202, in get_subj_alt_name
        except (x509.DuplicateExtension, x509.UnsupportedExtension,
                AttributeError: 'module' object has no attribute 'UnsupportedExtension'

Change-Id: I444dde829c91defb475e045aea094d74fc43e75b
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>

Merge "dump CI failures with kpartx and no /dev/loop devices"

Added scale-up documentation

Change-Id: I65770a4a879d890c75a2e5774579794fb9b591f6
Signed-off-by: Chornyi, TarasX <tarasx.chornyi@intel.com>

Merge "Make Sample VNF hugepages size configurable"

dump CI failures with kpartx and no /dev/loop devices

CI is failing due to unable to find loop device for kpartx

"cmd": [
    "kpartx",
    "-l",
    "/tmp/workspace/yardstick/yardstick-xenial-server.raw"
]

"stderr": "mount: could not find any device /dev/loop#Bad address\ncan't set up loop",

This error occurs when kpartx can't find any loop devices to use

https://build.opnfv.org/ci/job/yardstick-compass-virtual-daily-master/3261/console

JIRA: YARDSTICK-1054
JIRA: YARDSTICK-1096

Change-Id: Ib6131ce29c4f9e81386eb5471dd6107825798620
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Merge "update ci pod in the test suite according to Fraser"

Merge "Bugfix: Error when using API(v1) to run test suite"

Add k8-nosdn-stor4nfv-ha/noha scenario

JIRA: YARDSTICK-1090

We have k8-nosdn-stor4nfv-ha/noha scenario in compass,
so we need to add this files to trigger it.

Change-Id: I79709c53b8542434f7324ad907fa95b4855839d3
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>

Bugfix: Error when using API(v1) to run test suite

JIRA: YARDSTICK-1098

The reason is missing TasksHandler parameter to run test suite.

Change-Id: I9dd45caa87d0e39afbf7485443a6e566317f5cea
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>

update ci pod in the test suite according to Fraser

Change-Id: I999b44cc4e0ec1029c6efca224e691298a007689
Signed-off-by: rexlee8776 <limingjiang@huawei.com>

Merge "fix Dummy test context name"

Merge "Running Test Case in Latest Yardstick Docker Image shows Error"

Merge "Test case description and configuration file for yardstick_tc091:Control Node Openstack Service High Availability - Heat Api"

fix Dummy test context name

more dummy test fix base on I0ccb7e9fabdf6bc2890d2e4763f53baee06c87b2

Since I1447fb5ed447691eaeb0a97f928c0b3333799d07, the context name is
a mandatory parameter for every context.

JIRA: YARDSTICK-886

Change-Id: I10ee6bcc0507fa90b6e99261a98a96655fc66947
Signed-off-by: rexlee8776 <limingjiang@huawei.com>

Merge "Add k8-nosdn-stor4nfv-ha/noha scenario"

Install RabitMQ for RPC messaging between processes

In NetworkServices Tescases, the TGs (traffic generators) run the traffic in
a separate process. In order to synchronize the traffic injection and the
runner interval loops, an RPC server is needed to publish/subscribe events.

RabbitMQ is a well supported MQ in Linux (used in OpenStack or collectd)
and supported by Python implemented projects like oslo.messaging [1].

RabbitMQ default configuration:
- Port: 5672
- User/password: yardstick/yardstick

[1]https://github.com/openstack/oslo.messaging

JIRA: YARDSTICK-1068

Change-Id: I15db294ee430fb38e574a59b9ce1bf0f8b651a67
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Add "Scenario" class wait methods for runners

"Duration" and "Iteration" runners execute a passive wait during the
execution of the work process. This wait time is done at the end of
the scenario "run" method execution.

This patch adds a pre-run and post-run wait period, which will
depends on the Scenario executed. The wait time will be always the
same, but the execution order (pre-wait time, run method, post-wait
time) will depends on the Scenario.

By default, any Scenario will execute the "run" method and them will
wait the specified time.

NetworkServicesTestCase Scenario will wait the specified time and them
will execute the "run" method to retrieve the KPIs.

JIRA: YARDSTICK-1067

Change-Id: I6ad6bfc6978815b6b2d4df63f2ac2f8815fb5b8a
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Make "Scenario" class abstract

All scenario child classes must implement "run" method.

JIRA: YARDSTICK-1065

Change-Id: I35b78e380620967b49cd8cd23777a1aee6dfd140
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Improve "Libvirt.virsh_destroy_vm" function

Read the command exit code and log a warning in case the VM destroy
process went wrong.

JIRA: YARDSTICK-943

Change-Id: I2750b8d4a8f67af081c1988510cf5aca848a2cf1
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Improve NSB Standalone XML generation

Delayed the generation of the XML file until the last step. The
following functions will return a XML string insted:
- Libvirt.build_vm_xml
- SriovContext._enable_interfaces
- OvsDpdkContext._enable_interfaces

The XML file will be written just before copying the file to the
compute hosting the VMs.

JIRA: YARDSTICK-939

Change-Id: Icc80f4741903bbe335db4ebccab395b72fa87e82
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Improve OVS-DPDK setup process

Improved OVS-DPDK setup process:
- Added log string for each command executed.
- Removed unnecessary 'kill' command.
- Check if commands are executed correctly.

JIRA: YARDSTICK-922

Change-Id: I09fba96ffa39cde1e000a3c4382d474219c99985
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Improve "Libvirt.virsh_create_vm" function

Read the command exit code and raise an exception in case the VM boot
process went wrong.

JIRA: YARDSTICK-941

Change-Id: Ibabaae8983213a33799ad00fa4d9541bbabea857
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Improve OVS-DPDK boot process

Check if hugepages are present in the system.
Check if every command executed during the boot up process
exits correctly.

JIRA: YARDSTICK-917

Change-Id: Ic0c8a72199ad80e3a65488cc88d0692863751284
Signed-off-by: Rodolfo Alonso Hernandez <rodolfo.alonso.hernandez@intel.com>

Merge "Offline support for bottlenecks project"

Merge "Enforce pip version during installation"

Merge "Replace neutron router creation with shade."

TC042 bug fix

JIRA:YARDSTICK-786

yardstick DPDK image have problems to build.
We will fix this by using nsb image
the pktgen DPDK testcase will use sampleVNF image

Change-Id: I24b7b6a9acfe21ec956fa9c4ee7ebe7ae1787dbb
Signed-off-by: liyin <liyin11@huawei.com>

Offline support for bottlenecks project

JIRA: YARDSTICK-1053

Those three yaml files are for bottlenecks support offiline flavor
The ssh time out problem will fix in changing the duration time

Change-Id: Ifd317404a820d0446ffab5070211e48def1a0d2f
Signed-off-by: liyin <liyin11@huawei.com>

Add k8-nosdn-stor4nfv-ha/noha scenario

JIRA: YARDSTICK-1090

We have k8-nosdn-stor4nfv-ha/noha scenario in compass,
so we need to add this files to trigger it.

Change-Id: Iefa3a83407bff7d3c1de1959f35b6d8e5eecd145
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>

Bugfix: update test case base on node parameters in ci

JIRA: YARDSTICK-1082

Change-Id: I00080d0317a54b73a6cd571198e9446962c11b5a
Signed-off-by: rexlee8776 <limingjiang@huawei.com>

Merge "Fix for datasource field value"

Merge "Fix retrieving "options" section in "scenario""

Merge "Add "YARDSTICK_REPO_DIR" into tests path"

Merge "Modify the Docker diff file for ARM64"