increase number of open files for collectd

for collectd KPIs we need may need to open 10,000+ files

one file for each core * number of CPU counters we read.

Use the ansible pam_limits moules to increase the nofile
setting.  This should allow the VNF to boot with increased
number of files open

Change-Id: I46b75d5ca59e8e359c985a1e37daaee89897884c
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

task: drain background runner queues

We were not draining the queues in the background
runners.

Modify the main runner_join loop to loop over
all the runners and drain them.

The runner join method does extra work for periodic actions
so we can't change its behavior.

Instead create a new poll() method and use that
to check runner status

Change-Id: I9466ba40a6a4c45c82cedff279cbb4817c6b66ad
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Bug in  NSB Prox ACL Test 4 Port

JIRA: YARDSTICK-802

Generator was incorrectly configured to generated traffic for p0, p1 (twice), and p3

Should be p0, p1, p2 and p3

Change-Id: Ib62a3c6523b1eda635b8c727c36e140fd480003e
Signed-off-by: Daniel Martin Buckley <daniel.m.buckley@intel.com>

Updating NSBperf to print right testcase

Change-Id: I7be7bc61624fa14eb552477bdc701a44d65e68b7
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

private-> uplink and public->downlink

changing private and public to new uplink/downlink name

Change-Id: I8c3f582ba5592418a397e2500e0b998c3eed54f2
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

Enable intel_pmu plugin in collectd

Change-Id: I62c6416e5c79d0b90880985bb5c076b65e34192e
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

ping: always save rtt data, influxdb ignore empty data

If the SLA was failing we were raising AssertionError and not storing
the rtt in the data dict.  This caused influxdb parse errors because the
data was empty.

Fixup influxdb to ignore records with no data, so we don't try to parse
no data.

Change the ping logic to always record the rtt result even if the SLA
was not met.

Also fixup ping logic in cases where ping does not return results.  If
SLA is defined use SLA * 10 otherwise use large float that doesn't break
the grafana scale too much, maybe 999999

JIRA: YARDSTICK-809

Change-Id: Id2d51216581644a80e8c7b9aa98919a766008adf
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
(cherry picked from commit af011b16787e8fd9fc6f918b1d1427dd1be562ec)

Bugfix: Add opnfv_os-nosdn-ovs_dpdk testsuite for Compass4nfv

Change-Id: I35e6471bb561eb773bd188b36b912969188955d2
Signed-off-by: JingLu5 <lvjing5@huawei.com>

Bugfix: Add opnfv_os-odl_l2-moon-ha_daily testsuite for Compass4nfv

Change-Id: I46c6b56e1293c3d594804d77a73c391922f03653
Signed-off-by: JingLu5 <lvjing5@huawei.com>

Change 'SUCCESS' to 'PASS' when report to MongoDB

JIRA: YARDSTICK-822

To keep consistent with other testing project, we need to
change 'SUCCESS' to 'PASS', 'FAILED' to 'FAIL'.

Change-Id: I392e4dc6dd2e028cd43cba0d1bf8ee04618326d7
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>
(cherry picked from commit f4680c872fbb34816d809c41852740f5cb960fd1)

Bugfix: Reference for the tool used in TC044 mentioned in the user guide is wrong

JIRA: YARDSTICK-821

Change-Id: I09b189008df4ec2e6db92cf24c45d8206ed2062f
Signed-off-by: JingLu5 <lvjing5@huawei.com>

Merge "Doc amendment: env prepare command" into stable/euphrates

Doc amendment: env prepare command

JIRA: YARDSTICK-637

Change-Id: I7e5e1afb7b5a7e97aa070c03cded3be285d97cc2
Signed-off-by: JingLu5 <lvjing5@huawei.com>
(cherry picked from commit d91ddc83daa7ac112c764e74621b48455d038e7f)

Use "protocol" parameters in iperf3 yaml and task file

JIRA: YARDSTICK-755

There is a history problem that iperf use udp to set a net protocol.
This code will change it to protocol.
so you could use 'tcp','udp' and other protocol.

Change-Id: I1a101013dfe58165a3ed08aa77f0aa2f73d57a12
Signed-off-by: Ace Lee <liyin11@huawei.com>
(cherry picked from commit 10f85b332c4b1f55e651aeb9c45b328e1ebdc2af)

TC008 Interrupt Affinity Fix
The smp_affinity takes hex mask, currently it is given hex number.

Change-Id: Id738bfe10551dad1f5df48910a230b12f50d1b3d
JIRA: YARDSTICK-818
Signed-off-by: Jing Zhang <jing.c.zhang@nokia.com>

Bugfix: Update HW lb config VNF thread to start from core 1

Change-Id: Ib0007fa3689b90346598baaef5c6e88d4fdf444a
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

install pmu tools collectd plugin

Added review dependency, removed collectd commit-id
replicating bash script, added pmu roles to playbook

Create local mirror for event list generation, copy script
to image, and run event list generation at runtime if
intel_pmu is enabled.

Change-Id: Ie46a2b197f4d2037cf3eed194764ce9eb5670415
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

NSB Prox vPE Test

JIRA: YARDSTICK-802

Addition of Prox vPE test case
- The tests supports BM, Openstack Heat
- Supports 4 ports
- Grafana dashboards included
- Added support for parameters.lua
  for prox additional files
- Unit tests for code coverage

Change-Id: I5cccb351dacba88a293ae4b8aba1f0a803d62e6d
Signed-off-by: Abhijit Sinha <abhijit.sinha@intel.com>
Signed-off-by: Daniel MArtin Buckley <daniel.m.buckley@intel.com>
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

KVMFORNFV:Livemigration testcase changes in Yardstick

This patch includes changing the function call to place the
respective bash file to the given TARGET_SCRIPT path.

Change-Id: Iba2adf4bdb6d9045493f5099e29ade4e8e6b5cd3
Signed-off-by: Navya <navyax.bathula@intel.com>

libyajl: install from pkg

libyajl needed for ovs_events and ovs_stats, removed installing
from src and rather installing as deb pkg

Change-Id: Ic1ebf88304bbc7f4fa7b9b51c95f1e8e729e71e3
Signed-off-by: Dhaval Patel <dhaval.r.patel@intel.com>

Update load_images.sh to MCP@aarch64

The load_images.sh script still contained old functionality.
The changes are:
- don't use TestVM anymore, MCP does not load that by default like Fuel did
  upload the cirros image instead
- don't upload the kernel separately for aarch64, one part image works now
- add the needed extra-properties accordingly to each image

The script was tested on arm, the test are running now without errors

JIRA: ARMBAND-285

Change-Id: Iee8ce6cde43bebceb2edbb1310ea81c665876aff
Signed-off-by: Cristina Pauna <cristina.pauna@enea.com>
(cherry picked from commit af71c468fe83a5149e0e18cdb233e059302209e3)

fix random test_task unittest fail due to os.environ mock

sometime Jenkins fails due to what I guess are concurrency problems
in os.environ mock

======================================================================
FAIL: tests.unit.benchmark.core.test_task.TaskTestCase.test_parse_suite_with_constraint_no_args
tags: worker-10
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python3.5/unittest/mock.py", line 1157, in patched
    return func(*args, **keywargs)
  File "/home/jenkins/opnfv/slave_root/workspace/yardstick-verify-euphrates/tests/unit/benchmark/core/test_task.py", line 208, in test_parse_suite_with_constraint_no_args
    task_files, task_args, task_args_fnames = t.parse_suite()
  File "/home/jenkins/opnfv/slave_root/workspace/yardstick-verify-euphrates/yardstick/benchmark/core/task.py", line 455, in parse_suite
    cur_pod = os.environ.get('NODE_NAME', None)
  File "/usr/lib/python3.5/unittest/mock.py", line 917, in __call__
    return _mock_self._mock_call(*args, **kwargs)
  File "/usr/lib/python3.5/unittest/mock.py", line 976, in _mock_call
    result = next(effect)
StopIteration
Ran 1262 tests in 2.375s
FAILED (id=0, failures=1)
error: testr failed (1)
+ testr failing

Replace the mock decorator with a context manager to try to
reduce the scope and duration of the mock.

Change-Id: I342fe6c403e66c53ac4c39fd88fa9047cdfae5d9
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
(cherry picked from commit 2fadf7aec9e2761c39d29d8af1ee7d69d154652d)

Fixed compilation of latest samplevnf

added download_civetweb and install_civetweb roles.

Change-Id: Id4f60cc07b3befc7edb3cfc38773fe1a73a44c1f
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

add wait_for_instantiate

Change-Id: Idc54fc907dba4603984712fc43a0db8dfd4b7374
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

heat: add port name as port

Change-Id: Ibd159359c6f57d573a909d6841c121c15bf692c1
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Making sure yardstick container uses hosts timezone

This is required for influxdb results being reported
"in the future" when the timezones do not match.

Change-Id: Ic41e19d26c46b6ccfa6dacddb595236af19e437a
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

rewrite _generate_pod_yaml to combine name and pkey setting

just deepcopy and change.  This will probably break in
the future if we use anything other than dicts and list.

Change-Id: I9a9b0c5b09b3e3ebd7ed593bf6339ea030605f93
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Disable VNF deploy by default

Change-Id: I27bcc41c855f34fb1fd0332fc24e7bf0b2af4ec2
Signed-off-by: Deepak S <deepak.s@linux.intel.com>
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

NSB Prox minor refactor of BNG

JIRA: YARDSTICK-802

- Updated the BNG code with minor refactor.
- Corrected the CPE core name
- Update binsearch traffic profile with 64B

Change-Id: Iae0be766edb986520045655fa567651711813a8b
Signed-off-by: Abhijit Sinha <abhijit.sinha@intel.com>

Add intermediate variables for attacker,monitor,result_checker

JIRA: YARDSTICK-790

Change-Id: I6bb36c98b8673155d3142fc54cfb39315d5ce613
Signed-off-by: qiujuan <juan_qiu@tongji.edu.cn>

Added example TC with collectd on compute node

To be used with yardstick/etc/yardstick/nodes/pod.yaml.collectd.sample

Change-Id: I6eff4f6adf57596e06c685ab87b83699696ad7b6
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

Adding sample testcases to run on standalone context

- vFW
- vCGNAPT
- vACL
- UDP Replay
- vPE (Only OVS supported)

Change-Id: Idbc4d1d6bc1283e40d2fcb9457a871a9198ad147
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

Adding auto generate scale-out support for correlated traffic

Change-Id: I2755b596068545c1a3a672ceff47d814a44ae050
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

Adding 2 node ixia generic scale-out test case generation

Change-Id: I7da2d5bcd7c58c669e28a7271e4c6848c003e84a
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

Adding scale out templates for ovs_dpdk/sriov using 2 node setup

2 - node setup:
  - Traffic generator starts new stream on both uplink and downlink

This patch addes amsible scripts to enable scale_out testcases

- vfw

Change-Id: I0340636bce3e74cd6175f728b9e7e014a4eb2fd5
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

NSB PROX test hang fixes

The PROX tests were hanging in the duration
runner.

These are fixes for various errors:

raise error in collect_kpi if VNF is down

move prox dpdk_rebind after collectd stop
fix dpdk nicbind rebind to group by drivers

prox: raise error in collect_kpi if the VNF is down
prox: add VNF_TYPE for consistency

sample_vnf: debug and fix kill_vnf
pkill is not matching some executable names,
add some debug process dumps and try switching
back to killall until we can find the issue

sample_vnf: add default timeout, so we can override
default 3600 SSH timeout

collect_kpi is the point at which we check
the VNFs and TGs for failures or exits

queues are the problem make sure we aren't silently blocking on
non-empty queues by canceling join thread in subprocess

fixup duration runner to close queues
and other attempt to stop duration runner
from hanging

VnfdHelper: memoize port_num

resource: fail if ssh can't connect
at the end of 3600 second test our ssh connection
is dead, so we can't actually stop collectd
unless we reconnect

fix stop() logic to ignore ssh errors

Change-Id: I6c8e682a80cb9d00362e2fef4a46df080f304e55
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Switched nsb_setup to pull stable yardstick docker image

Change-Id: I4474ee65899d5f60e657835e8d80de0d372abd22
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

Enable multi-port perf in trex as TG

set TRex -c option for threads per port based on
hardware number of queues.

We can't auto-detect number of queues and we can't
use more than one thread per core on systems with single-queue
interfaces, so move the option to the config file

options:
  tg_0:
    queues_per_port: 2

also enable trex debug by removing >/dev/null redirection

options:
  tg_0:
    trex_server_debug: true

Change-Id: I46da187849282bf28f4ef5b333e1ae890e202768
Signed-off-by: Deepak S <deepak.s@linux.intel.com>
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

runners: add timeout to queue put

we don't want to block the test waiting to put KPIs
Add moderate timeout.  In case we do timeout, it
doesn't matter if we drop intermitten KPIs

Change-Id: I049c785355993e6b286748a5c897d54dd2923dc9
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

add collectd resource node capability

allow manually adding collectd nodes using Node context.

if a node is present with a collectd config dict then
we can create a ResourceProfile object for it
and connect to collectd.

example

nodes:
-
    name: compute_0
    role: Compute
    ip: 1.1.1.1
    user: root
    password: r00t
    collectd:
        interval: 5
        plugins:
            ovs_stats: {}

Change-Id: Ie0c00fdb58373206071daa1fb13faf175c4313e0
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Optimize Core usage for the sampleVNFs

Change-Id: I80aa7e796b9ca4c4881c78310860e293a4a75560
Signed-off-by: Deepak S <deepak.s@linux.intel.com>
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Prox L3FWD 2-port test fix

JIRA: YARDSTICK-802

Updated the handle config for l3fwd 2 port test
The tx and rx descripters is removed as they
were not there in original DATS config.
The BM test was dropping packets beacuse of this.

Change-Id: I40d113267cbb3376a772b5a5aaecf74bea9d06fb
Signed-off-by: Abhijit Sinha <abhijit.sinha@intel.com>

drain runner queue and undo cancel_join_thread

Sometimes the runners can hang.  Initially
debugging lead to the queue join thread, so I thought
we could cancel all the join threads and everything would be okay.

But it turns out canceling the queue join threads can lead
to corruption of the queues, so when we go to drain the queues
the task hangs.

But it also turns out that we were not properly draining
the queues in the task process.  We were waiting for all
the runners to exit, then draining the queues.

This is bad and will cause the queues to fill up and hang
and/or drop data or corrupt the queues.

The proper fix seems to be to draining the queues in a
loop before calling join with a timeout.

Also modified the queue drain loops to no block on queue.get()

Revert "cancel all queue join threads"

This reverts commit 75c0e3a54b8f6e8fd77c7d9d95decab830159929.

Revert "duration runner: add teardown and cancel all queue join threads"

This reverts commit 7eb6abb6931b24e085b139cc3500f4497cdde57d.

Change-Id: Ic4f8e814cf23615621c1250535967716b425ac18
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

exec_tests: remove releng clone code

we bind-mount the credentials into the container
so we don't need to clone releng

Change-Id: I838745a1470015a693f1801c93b930e8df06ef4b
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Ensure that at least one handler is available

https://jira.opnfv.org/browse/YARDSTICK-773?filter=-3

Remove dependency of yardstick on utils methods

Change-Id: Iadf502364a7f08c279a8f0d17d7e45e8047f4066
Signed-off-by: Edward MacGillivray <edward.s.macgillivray@intel.com>

ansible: add grub.cfg hugepages regexp replace

when building the image in chroot we can't run
update-grub of grub-mkconfig because both commands
probe for root partitions, which doesn't work
in the chroot.

Instead we manually regexp replace the kernel boot
line in /boot/**/grub*.cfg

Change-Id: I31fac2a3db4d1d044f11357b1715119066509676
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

code inspection fixes: test_pktgen

Change-Id: I05cb069984b7674924cfcb1ed023048c0aa0c444
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Adding example testcase to enable multiport support for http

Change-Id: I780aa3ea6b04df08baffb5ee5beff66bdc37f37e
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

Fix src/dst ip update in ixia configuration

Change-Id: I03976c182e164a512a5ef48ad20928f27451c3f6
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

Enabling multi_VM & multi port launch in standalone context

new context names:
- SRIOV - StandaloneSriov
- OvsDpdk - StandaloneOvsDpdk

- Seperate helper, libvirt, server info class
- Allow multi-port and multi-VM support.

Change-Id: I3c65e4535082fa0e2f4c6ee11c3bca9ccfdc01b8
Signed-off-by: Deepak S <deepak.s@linux.intel.com>
Signed-off-by: Martin Banszel <martinx.banszel@intel.com>

Grafana Dashboard update for Prox Tests

JIRA: YARDSTICK-802

Updated the Grafana dashboards to collect KPI's
correctly & incorporate scale up changes in topology

Influx-db stores the KPI’s fields as: <TG><Port><kpi>
TG has changed from  : tg__1 to tg__0
VNF has changed from : vnf__1 to vnf__0

Change-Id: I3bd8be61210906bc4de95c9b1986033572d93277
Signed-off-by: Abhijit Sinha <abhijit.sinha@intel.com>

Merge "load_images.sh: Add missing ssh_options" into stable/euphrates

load_images.sh: Add missing ssh_options

Empty ssh_options led to ssh login failure, which stopped the whole
script.

Change-Id: I8374a30a02b14d04eb0f623a0c58d7ebed77a589
Signed-off-by: Alexandru Avadanii <Alexandru.Avadanii@enea.com>
(cherry picked from commit 77c7d358bc6a61e2a5c24c6056cc3e40561f27c7)

CI: add new test cases into CI so they can have result

Change-Id: Ib2b165c41121b29b2af378dc420c2360786de3f0
Signed-off-by: rexlee8776 <limingjiang@huawei.com>
(cherry picked from commit 4aecf8839e12260284603266c4b93dc72b33ea36)

Auto create ixia config based on the traffic profile

Change-Id: I031cc7f24f0c0816eb577a4d1606a714f68a5f83
Signed-off-by: Deepak S <deepak.s@linux.intel.com>
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

replace ansible modules

Change-Id: Ia7c1ce781075142910a6c618a9a23f34a710dfe9
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Adapt getting the image architecture to mcp

The part which gets the image architecture still uses the old fuel command.
Adapted the script to use the mcp command.

JIRA: ARMBAND-285

Change-Id: Iddce77c3a856ea11f9994861151db89a402e735e
Signed-off-by: Cristina Pauna <cristina.pauna@enea.com>

dpdk: pciutils and lspci is required for dpdk-devbind.py

Change-Id: I8529f33d849fc659b557e1d19761c0765ebda57e
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Adjusted IPs in some prox tests.

Change-Id: Iff39fa5b90d1ae5534ce1fff31dfb84c1e47340c
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

Switched to stable/euphrates for samplevnfs projext

Change-Id: I0305c263740c25bdfa3bb2667838ca2b75d33ba0
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

add collectd ansible

required DPDK fixes to enable building collectd with dpdkstats
need /usr/lib/dpdk-pmd symlinks

Compiling all this code requies more space than the default ubuntu
cloud-image, so we need to resize the cloud image somehow

Change-Id: Ib39fc94ba8101ca1aebc7d2a710be88b0338f7c3
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

Added hw_vif_multiqueue_enabled metadata to the added samplevnfs image in OpenStack.

Change-Id: Iad1dce6b0d0a1c010aa62f2fdf54c5d0560ed0c7
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

vnf_generic: Fix str object has no attribute items

When an IP range is specified in src_ip/dst_ip like:
src_ip:
   - '152.16.100.180-152.16.100.181'
yardstick would return str object has no attribute items error.
This change will return the IP range as is if type is str.

Change-Id: I3b097777f0d85b0600207157bebba18987ea2275
Signed-off-by: Dino Simeon Madarang <dinox.madarang@intel.com>
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

NSB: ignore VNF node if there is no VNF model

Change-Id: I571289c744441a6114c6d45ac6cbd7cfc0dafef2
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

collectd: add config file templates as package_data

needed for pkg_resources.resource_string

Change-Id: Ief7a8fcd88fd6d371851a26e1559351963979e68
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Addition of Prox NSB BNG and BNG-QoS test

JIRA: YARDSTICK-802

Added Prox BNG and BNG-QoS Test
- The tests supports BM, Openstack Heat
- Supports 4 ports
- Test added for BNG traffic profile
- Fixed the Prox heat test cases with
  proper upstream and downstream links
- Grafana Dashboard for BNG & BNG-QoS added
- Increased the test Duration to 300

TODO:
- Test does not Terminate correctly

Update:
Added new helper class for run_test: Genric, MPLS
and BNG tests.

Change-Id: Ib40811bedb45a3c3030643943f32679a4044e076
Signed-off-by: Abhijit Sinha <abhijit.sinha@intel.com>
Signed-off-by: Edward MacGillivray <edward.s.macgillivray@intel.com>
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Daniel Martin Buckley <daniel.m.buckley@intel.com>

Terminate the run traffic if initiated after traffic is closed

Change-Id: I89fc4bccb5b5a0d49d804fca72b59e53ff77a805
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

collectd: write config file from Jinja2 template

We have the collectd.conf inside the python package
so instead of copying it from various places,
write the template directly to the remote system.

collectd: read collect.conf template with pkgresources

read the collectd.conf file as a string directly
and upload without creating temp file

use Jinja2 template, disable failing plugins

use proper Jinja2 template, disable the plugins that
were failing to load and blocking startup

add support for per-testcase collectd.conf config
using YAML

add support for custom interval, default is 25 seconds

Change-Id: Id904f7b7c9f41a9dd7adf5dfa06c064d65c25d2d
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Adding multi-port support for ixia taffic generator

Change-Id: Ic8aa130f3cdc7bd8dec39d06a6b824340bf658b2
Signed-off-by: Deepak S <deepak.s@linux.intel.com>
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

bugfix: tc055 fail due to options with no value

JIRA: YARDSTICK-816

Change-Id: Ib7eb411b940775915c6c9f87ac5cdc9825069467
Signed-off-by: rexlee8776 <limingjiang@huawei.com>

Fix vPE VNF characterization issues.

Change-Id: Ia934128777d2839f6d2b940857c266fc3e2bd4a1
Signed-off-by: Deepak S <deepak.s@linux.intel.com>

[PATCH] Patch for Yardstick arm64 netperf_install.bash

Change-Id: Ic384c4603e7482f150fd4c4d2d6a6448a45ddb9e
Signed-off-by: Trevor Tao <trevor.tao@arm.com>

Add nodeSelector to enable selecting the desired
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>

Enable IP_ROUTING for netperf UDP_STREAM test

To avoid the following ERROR when netperf test between
2 different subnets:
ERROR netperf: send_omni: send_data failed: Network is unreachable

For detail, please see:
https://serverfault.com/questions/802320/netperf-iptables-\
masquerade-network-unreachable
Or:
https://stackoverflow.com/questions/11981480/error-in-running-\
netperf-udp-stream-over-openvpn

Change-Id: I62b202844861440deaf3bf0f65b41561bd87ae87
Signed-off-by: Trevor Tao <trevor.tao@arm.com>

Read user & password from yardstick.conf in Grafana configuration

JIRA: YARDSTICK-812

Currently grafana data source configuration is hardcoding .
It is a risk.
so I read it from yardstick.conf.

Change-Id: I8a9c8afbce6c4534fc43a0bfb5c56d67a8b59db0
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>

Merge "Get node IPs and IDs according to env" into stable/euphrates

bugfix: tc054 default value is wrong

management route ip is not common in most SUTs, so it should
be removed.
also huawei pod1 ipmi info should updated so this test case
can be added into ci later

Change-Id: I3a29c59c473ee7087d4d61753ffc955b061571fb
Signed-off-by: rexlee8776 <limingjiang@huawei.com>
(cherry picked from commit 8701e63e3daf508d8e3482f0a344554d17ff6e24)

Remove checkno.png and checkyes.png due to license issue

JIRA: YARDSTICK-817

Since checkno.png and checkyes.png is not Apache-2 license based.
so we need to remove them.

Change-Id: I40dd303fb54a3736ca969ac1c186d2cd23408436
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>
(cherry picked from commit da0163b7b7aaf3ede4e757a0b7d94a5ea99b1083)

Get node IPs and IDs according to env

Extract node IPs and IDs for each node having a controller or compute
role (name starting with the "cmp" or "ctl" prefix) and add them into
the $pod_yaml file, analogous to the previous implementation.
Since node IDs are expected to be unique and integers (condition that
does not match the salt node ID format), they are substituted with an
incremented index for each controller/compute node in the environment.

Change-Id: Id90626edc3f098bd96343336b2be179721dee5a1
Signed-off-by: Catalina Focsa <catalina.focsa@enea.com>
(cherry picked from commit 6892687967d2d5ac8db37dd67b3e52d9f775eda6)

Create test suite for Euphrates release

JIRA: YARDSTICK-814

Test suite:
"opnfv_os-odl-fdio-ha_daily.yaml",
"opnfv_os-odl-dvr-noha_daily.yaml",
"opnfv_os-odl-sfc-noha_daily.yaml"
are missing in yardstick-apex-baremetal-daily-euphrates job.
We need to create them.

Change-Id: I6d8bbeb17cd887776f1f3b401ec80523ea90d3c1
Signed-off-by: chenjiankun <chenjiankun1@huawei.com>
(cherry picked from commit 6ecb9a6d50345277645633b1bed4d255dc434222)

task: fixup s/exe_info/exc_info

JIRA: YARDSTICK-785

Change-Id: Ib37498e8df6a520f1d03256b73346fcedab3a177
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
(cherry picked from commit 9ea225f671b774c6e373dbaab146d68cac16194e)

tg_rfc2544_ixia: log exceptions and fix list iter

Change-Id: Ia9722604b7c8ae23e784e780f113d012de544d4b
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Add service in 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>

bugfix: tc025 should use ipmi to poweroff

if it use shutdown, it'll take several minutes to shutdown,
leads to the ipmi power on command fails

Change-Id: I74b61325cbcc3a6ec070d2fa103accf84f29b0fa
Signed-off-by: root <limingjiang@huawei.com>

Introduced timeout to post method of HttpClient

We seen cases where grafana container bring-up code would fail,
because of too quick access to the http api. Added 10sec timeout
for the first query of the API.

Change-Id: Ifc95a626d0ab5552a1f26fb167fc3f65791392d7
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

tg_rfc2544_ixia: use traffic id from JSON to update mac

we were using raw sort index of the interfaces to
set the MAC address, but we should be using the
traffic id from the static JSON instead.

Change-Id: I13284db04abb3eaf8c9826974a9e5aa1c37b3891
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Bumped samplevnfs flavor disk size to 6G.

Since we increased the images size the 4G is not sufficient anymore.

Change-Id: Iae25cf4cfb7a6cc69c8d28771c183a2342ac38d0
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

ixia: renamed execute to execute_traffic

ixia: pass ports to generate_samples

Change-Id: I90d12fa2ce8cd4d1c2a18bdcf70027f6d9e3f77f
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

nsb_setup.sh: update to Ansible 2.3.2

Change-Id: I2f700fbb169d02d126fe7ea22721bebf127c1206
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

cgnapt: mock time.sleep in unitests

Change-Id: I26957977e6dcd0392078a543a6907a550711c702
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

requirements: bump Ansible version to 2.3

Change-Id: Ib429ba24d2b7287b6ec4e749386da0e1242d6a20
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Tweaked nsb_setup to always re-download latest container

and re-create the container.

Change-Id: I21204ddf97e2cccc2d5a762f5d910068bda1a948
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

prox: fix TG KPIs

The problem is that we share the same ProxResourceHelper
for both VNF and TG.

For VNF we want to talk to resource.py and get collectd KPIs.
For TG we need to read from the queue the TG calculated KPIs and
we also want collectd KPIs.

workaround is to use a different method name collect_collectd_kpi
for VNFs

Change-Id: Icc2132758e37ce210f5600a0cd433077930208e5
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

Addition of PROX NSB tests to yardstick

JIRA: YARDSTICK-802

Addition of PROX L2FWD_Multiflow, ACL, Load Balancing  plus
grafana dashboards

Supports 2 and 4 port Baremetal & Heat

Change-Id: I1f3990d5451de265ee3901302569c355ece3b146
Signed-off-by: Daniel Martin Buckley <daniel.m.buckley@intel.com>

prox: use find_relative_file when uploading

the prox files were being found correctly.
if we use find_relative_file they will lookup
relative to the task_path

Change-Id: Ifde5d07df5ccfbfeba015b2f43bd8b53e89a00b7
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

add opnfv_os-odl-fdio-dvr-ha_daily.yaml scenario

Change-Id: I723477edf810a220816a2e67aa80f7f144efb3a6
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

prox: pass prox_config_dict between Processes using queue

we generate the prox_config_dict in the _run Process,
but we also need it in the _traffic_runner Process to
get core info.

use a queue to pass the config list between the processes

enable collect_kpi

Change-Id: Ibaf41d606e559a87addf43d6ddaed206dbd2d20c
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Edward MacGillivray <edward.s.macgillivray@intel.com>

Fixed few gaps when running on a minimal version of system

- we need to make sure we have lsof installed
- we need to update cache inside the image, because we are unable to install some of them

Change-Id: Ic555489779e9096540001cb9c62ea2ab25c1ae90
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>

prox testcases: private -> uplink,public -> downlink, vnf_0, tg_0

Change-Id: I85afff4582bf538fcd0be5b4db1405a4da2573f9
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>

UDP_Replay: enable collectd KPIs

Change-Id: I92146411707a9ec29864d164dbd63b96d05bffe0
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>