Ping monitor failure in tacker dsvm gate

[Sridhar Ramaswamy]

Tacker dsvm functional test for ping monitor is consistently failing,

tacker.tests.functional.vnfm.test_vnf_monitoring.VnfTestPingMonitor.test_create_delete_vnf_monitoring_tosca_template

.. with,

2016-08-26 13:51:24.063027 | 2016-08-26 13:51:24.062 | File "/opt/stack/new/tacker/.tox/dsvm-functional/local/lib/python2.7/site-packages/testtools/testcase.py", line 498, in assertThat
2016-08-26 13:51:24.066164 | 2016-08-26 13:51:24.065 | raise mismatch_error
2016-08-26 13:51:24.067495 | 2016-08-26 13:51:24.067 | testtools.matchers._impl.MismatchError: 'DEAD' != None

More details available in,

http://paste.openstack.org/show/564534/
http://logs.openstack.org/05/352205/11/check/gate-tacker-dsvm-functional

Triage shows the ping never stops working, that is 'ifconfig down eth0' in the user-data doesn't take effect. When ping monitor never fails to trigger the DEAD -> ACTIVE respawn action that the test script is trying to validate.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/363398

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/363398
Committed: https://git.openstack.org/cgit/openstack/tacker/commit/?id=b1896e743ea4ed2081fc7863c09955fede1200d9
Submitter: Jenkins
Branch: master

commit b1896e743ea4ed2081fc7863c09955fede1200d9
Author: gong yong sheng <[e-mail]>
Date: Wed Aug 31 13:08:54 2016 +0800

    Fix the monitor bug

    The VDU VM's dead simulation depends on metadata service which needs
    virtual routers, or config drive.

    This patch adds "config_drive: true" to all templates which need user
    data.

    In addition, some verbose debug information is inserted into code to
    help to diagnose problem.

    Change-Id: I066ce769de17ad3179193b436f5a644b76233a1e
    Closes-bug: #1617923

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/tacker 0.5.0 release.