Age | Commit message (Collapse) | Author | Files | Lines |
|
JIRA: YARDSTICK-255
add RAMspeed scenario for measuring memory bandwidth.
Change-Id: Iba740d7fdb394d96f32ee824925bbbf8fb689614
Signed-off-by: JingLu5 <lvjing5@huawei.com>
|
|
|
|
Using LMBench to measure latency of cache.
two parameter can be configured (repetition and warmup)
Change-Id: I5e4ecca0f9dd9c9ce2cecce3623dd8347ab2b5b1
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
Change-Id: I13cc3f6cdfc243391df6e15827428352a8d6634d
Signed-off-by: rexlee8776 <limingjiang@huawei.com>
|
|
This change adds missing license information.
Change-Id: I2ddc85e3fb7b92a86b8a0f054d4ee177af198c8f
Signed-off-by: Ana C <ana.cunha@ericsson.com>
|
|
Change-Id: Ia3760a4fcc17f810940240d1778939333e8ffdc6
Signed-off-by: wym_libra <yimin.wang@huawei.com>
|
|
As we known, the network of os+odl scenario is still not available.
so this patch still need to do more debug work.
however, Jan 19th is code freeze day, i think a good way is to merge it at first
when os+odl become available, i will do rest debug work and fix it
Change-Id: I518f759069e74d2e57e40b52dfecaef4b4d26a02
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
Change-Id: I6587794b7206cb3d5f13b6c448670d1c97ecd596
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
JIRA:YARDSTICK-187
Change-Id: Ia15d17afdef145f7b230a8a4d25a61eed5cdfd76
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
In Heat Liberty release OS::Nova::Server will always use the user
pre-configured in the image (e.g. "fedora" for stock Fedora cloud
images, "ubuntu" for stock Ubuntu cloud images, "cloud-user" for
stock CentOS cloud images etc)
Change all ec2-user -> ubuntu
Add admin-user in Heat model for backwards compatibility.
Refer below links for detalis:
https://etherpad.opnfv.org/p/yardstick_release_b_troubleshooting
https://github.com/openstack/heat/commit/e423bec7f10b0f5d07f05d195b3b7860f6bceb00
http://blog.scottlowe.org/2015/04/23/ubuntu-openstack-heat-cloud-init/
JIRA: -
Change-Id: I6b8b2b21daf113a3a86aee1126b0c3e74737ef4f
Signed-off-by: QiLiang <liangqi1@huawei.com>
|
|
|
|
JIRA: YARDSTICK-122
Change-Id: I8144215059a9abea08314a4c1e6a733dcdf0df53
Signed-off-by: QiLiang <liangqi1@huawei.com>
|
|
1) add "attacker_baremetal.py" for fault injection
2) modify the monitor to excute on remote node after ssh connection
3) move all shell scripts together
JIRA: YARDSTICK-182
Change-Id: Ibb9dc908224ddb8b99a0140b75c1a046503f6dfb
Signed-off-by: wym_libra <yimin.wang@huawei.com>
|
|
idea: refact the Monitor class in old file "monitor.py" with the base
class and sub-class.
detail:
1) the BaseMonitor is the base class of other monitor
2) each monitor run in independent process
3) there are two monitor("openstack-cmd" and "process") for the first test case
4) MonitorMgr class used to manager monitor process
JIRA: YARDSTICK-149
Change-Id: I2eede94481f740812212e6cb673d175b5f543c15
Signed-off-by: wym_libra <yimin.wang@huawei.com>
|
|
with jnon and fatih's help, new docker image has been uploaded
so this part is about parser verify
validating output against expected outcome.
Change-Id: If50d241a5338888f14fd11a752dc72678e0c569b
JIRA:YARDSTICK-224
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
JIRA:YARDSTICK-187
Change-Id: I1cecd400b4449a09d22d43f4a42e889f00dd4fe7
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
Dummy Context Usage:
- if no context specified in the task file then automatically use
Dummy Context
- or specify the context with type Dummy in the task file, like
context:
type: Dummy
Note: context without type name default use Heat Context.
(e.g. samples/fio.yaml)
JIRA: -
Change-Id: I7f798a7260bdd6ac24902e2c835a3b121319fd8c
Signed-off-by: QiLiang <liangqi1@huawei.com>
|
|
JIRA:YARDSTICK-184
Change-Id: Iedd4a3708e08305b1c8fa7a8e1766ceef03ab8bb
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
Stop size parameter's type should be float to allow for performing
measurements on only L1 and L2 cache if needed.
JIRA: -
Change-Id: Ibc4d052c343e9dcccf6101c6c949172bb5a0f9e1
Signed-off-by: Kristian Hunt <kristian.hunt@gmail.com>
|
|
|
|
Add test case description for HA of OpenStack Controll Node abnormally
shutdown (TC025) and the sample .yaml task file.
** Please note that the SLA value has not yet been determined. **
JIRA: YARDSTICK-135
Change-Id: I70f1afb0c9cfa5446aac08b7b13e543a9ec20898
Signed-off-by: wym_libra <yimin.wang@huawei.com>
|
|
The Arithmetic runner is modified to accept several named
parameters to iterate over in each scenario. Possible to chose
either to iterate through them in a nested-for-loop fashion
(default) or in a tupled loop fashion (combine i:th element
of each input). Still possible to only provide one parameter
to step/iterate over, as before.
All existing yaml files with Arithmetic types have been modified
to run the tests as before, and to provide same results as before
this change. The only change is in sample file pktgen.yaml to
provide a valid example of how to use this modification.
It is also possible to step backwards, i.e. start > stop, with
a negative step parameter value.
(Noticed that tests are terminated when there is SLA error
problems in "monitor" mode. Should be fixed in another commit)
Change-Id: Ib2ebd24b71dd55d6817cee8e67026a0dd13a9e17
JIRA: -
|
|
refactor the attacker implement.
1) BaseAttacker is added
2) a simple attacker named "kill-process" inherit the BaseAttacker
3) serviceha.py selects an attacker through the BaseAttacker by attacker name
JIRA: YARDSTICK-149
Change-Id: Ib718d5edc6b5e14bc3ea0592e0146468ff70b43e
Signed-off-by: wym_libra <yimin.wang@huawei.com>
|
|
This change adds the possibility to run scenarios as "background
tasks".
Background scenarios/tasks:
- are started before all "normal scenarios"
- runs in parallel with "normal scenarios"
- terminates when all "normal scenarios" have completed
their tasks
They are intended as a way to perform background tasks, e.g. collect
data such as cpuload etc, in parallel with the execution of normal
benchmarking scenarios.
Note that we already have the 'run_in_parallel' attribute but
this attribute has a couple of issues and do not solve all the
uses cases.
Change-Id: I9c5230bfdbbb66030f57b658ce1db87ff2c2d62b
Signed-off-by: Jo¶rgen Karlsson <jorgen.w.karlsson@ericsson.com>
|
|
This scenario reads processor and system load statistics
and does not run any benchmark tests.
The scenario is intended to be run in parallell with
other scenarios in order to collect processor and system
load statistics.
System load is read from /proc/loadavg.
Processor usage stats is read using the 'mpstat' utility if
it has been installed on the host.
If 'mpstat' is not installed on the host processor usage stats
is read from /proc/stats.
Change-Id: I7156e0c941100023571db750de7540786a4fedb8
JIRA: YARDSTICK-181
Signed-off-by: Jo¶rgen Karlsson <jorgen.w.karlsson@ericsson.com>
|
|
1)stop an openstack service
2)then monitor the corresponding api and check the availability of it
3)recovery the openstack service
JIRA: YARDSTICK-149
Change-Id: Id7b77d2f5c71844729c04f37442c8cfaa270ab12
Signed-off-by: wym_libra <yimin.wang@huawei.com>
|
|
Initial NodeContext implementation to support BareMetal,
Controller, Compute scenarios.
Usage:
0) install yardstick
1) mkdir -p /etc/yardstick/nodes
2) cp <yardstick_repo>/etc/yardstick/nodes/pod.yaml.sample \
/etc/yardstick/nodes/pod.yaml
3) edit /etc/yardstick/nodes/pod.yaml (make show ip, username,
ssh key are configured correctly)
4) yardstick -d task start \
<yardstick_repo>/samples/ping-node-context.yaml
5) cat /tmp/yardstick.out
Design etherpad link:
https://etherpad.opnfv.org/p/yardstick_framework
JIRA: YARDSTICK-169
Change-Id: I3f6ade8243e68d88326f23ed213edb32c638ed32
Signed-off-by: QiLiang <liangqi1@huawei.com>
|
|
Lmbench scenario has now two scripts and will choose between them,
based on whether the intention is to run memory latency or bandwidth
tests. Added also unit test file for this scenario.
JIRA: YARDSTICK-113
Change-Id: I2ba4dbef31f3cafbdb3c583ece5ed9512a906896
Signed-off-by: Kristian Hunt <kristian.hunt@gmail.com>
|
|
Heat context code refactor to cater for the evolution of the
Yardstick framework.
At test_case.yaml context segment add "type" to indicate the
context type, see samples/ping-heat-context.yaml for an example.
And the default context type is Heat, so the existing yaml file
do not need to change.
JIRA: YARDSTICK-168
Change-Id: Ida0ce12c17cd9b88d7acfb4c9eb1ac6986394b38
Signed-off-by: QiLiang <liangqi1@huawei.com>
|
|
Support measuring operating system's high resolution by using Cyclictest.
JIRA: YARDSTICK-122
Change-Id: I2e00ce117e263deaaf52cd2d663b845bd5b65432
Signed-off-by: QiLiang <liangqi1@huawei.com>
|
|
As agreed in the IRC meeting, use iperf for YARDSTICK-112;
support for netperf shall also be included in the framework (Yardstick-121).
bulk data test and req/rsp test are supportted by netperf
JIRA:YARDSTICK-121
Change-Id: I0e934dc067802c2792d751a19a187cd00ad8741b
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
|
|
This patch proposes a more permanent solution to replace the already
existing one (YARDSTICK-106).
A test suite is a .yaml file which consists of multiple task files which
are meant to run as a one set. Each test case in the test suite has to
specify the location of the test .yaml task file and can specify additional
task arguments or a task arguments file.
It is possible to specify the directory where the task files are located
in the suite file. If it is left unspecified then it defaults to
tests/opnfv/test_cases according to https://etherpad.opnfv.org/p/yardstick_folders
A Command line argument --suite is added to the task command. This
argument indicates that the intention is to run a test suite.
JIRA: YARDSTICK-107
Change-Id: I69508b3b536dd592b6240a2a32dd5d665e0ac4b4
Signed-off-by: Kristian Hunt <kristian.hunt@gmail.com>
|
|
JIRA: YARDSTICK-34
Change-Id: I782ba5845f8bd54a19bad078fe7be546400f7524
Signed-off-by: houjingwen <houjingwen@huawei.com>
|
|
1)create test case("iperfs-jitter") for measuring packet delay variation
2)and modify "iperf3.py" to support it
JIRA:YARDSTICK-131
Change-Id: Ic3416a713b69dc7687f6f85ecc0b79da8dfb7ed9
Signed-off-by: wym_libra <yimin.wang@huawei.com>
|
|
This patch adds YAML document start boundaries (YAML spec 4.3.2)
to samples, where the boundaries are missing.
JIRA: -
Change-Id: I6e67d4d73a449ad901392c017220a61f642716ac
Signed-off-by: Kristian Hunt <kristian.hunt@gmail.com>
|
|
1) when duration>3600s in fio.yaml, ssh time out
2) sometimes the latency value in test result is null
3) update plot.py, fit for code update in fio.py
4) small bug in file.py (result output dump)
Add --output-format=json in default args, so fio command
can return json format data.
JIRA: YARDSTICK-143
Change-Id: Ie02977b8c9f11986a1eed66896b84d18db3d2211
Signed-off-by: houjingwen <houjingwen@huawei.com>
|
|
This patch will make the iperf sample test case consistent with other
test cases by using the yardstick-trusty-server image and making it
possible to run this test case "out of the box" without requiring the
modification of the sample file.
Change-Id: I9b0da0d94f9c33f07bca79067a1eb5d615e2c15a
JIRA: YARDSTICK-148
Signed-off-by: Kristian Hunt <kristian.hunt@gmail.com>
|
|
Example invocation:
$ export EXTERNAL_NETWORK=net04_ext
$ yardstick task start samples/ping.yaml
The default external network is "net04_ext"
external_net should not be possible set in yaml anymore.
JIRA: YARDSTICK-54
Change-Id: I8c7b2850801f312c6878e9a363b06cdc8ade09bf
Signed-off-by: QiLiang <liangqi1@huawei.com>
|
|
Add support in task file for the template syntax based on Jinja2.
JIRA:YARDSTICK-101
Change-Id: I24be133ba590510612d97a1fce6c024e6edb57e4
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
FIO storage performance varification support variable iodepth.
JIRA: YARDSTICK-119
Change-Id: Ib6af55c5fbcaa7727a7541b13053b5e4ce8d3203
Signed-off-by: houjingwen <houjingwen@huawei.com>
|
|
A Iteration runner is roughly the same as a Duration runner but runs
for a configurable number of times/iterations instead of time.
Basically just the termination condition is different.
modify the unit and default value of iteration.
rename iteration to iterations
JIRA:YARDSTICK-49
Change-Id: I67f4014dc3cf923cd31cc2e990e2f7219bce40fe
Signed-off-by: kubi <jean.gaoliang@huawei.com>
|
|
Supports using perf software event counters. No hardware events are
currently supported. It is possible to add some load to the system
using the load parameter, otherwise the system will sleep while the
measurements are collected.
It is possible to configure SLA for any event measured.
Change-Id: Ic413d940093aadd10dc32888ea416aa94316a6fe
JIRA: YARDSTICK-51
Signed-off-by: Kristian Hunt <kristian.hunt@gmail.com>
|
|
Supports measuring memory read latency using Lmbench.
SLA can be verified for maximum latency time in nanoseconds
compared against results from every measurement taken.
Change-Id: Ia187e68eaa03341ba14b6e9eb986afb1d00dd0f6
JIRA: YARDSTICK-89
JIRA: YARDSTICK-92
Signed-off-by: Kristian Hunt <kristian.hunt@gmail.com>
|
|
By request, changed the misleading 'client' and 'server' designations
under context:servers in the sample files, to avoid misunderstandings
and convey the message that their names are irrelevant.
Uses the Greek pantheon as established in
https://gerrit.opnfv.org/gerrit/#/c/946/
JIRA: -
Change-Id: I28438340f7cbad679c2ffd4d147fbf0c7d2a379d
Signed-off-by: raindirve <seanw@kth.se>
|
|
pre-start and post-stop intention is to be used to gather
information about the target system.
single-shot and periodic-action intention is to perform
actions on the infrastructure or cloud resources. For example
server live migration or network interface down.
Example of what can be added in the runner section:
pre-start-action:
command: "heat stack-show demo"
periodic-action:
interval: 10
command: "ifconfig vboxnet1"
single-shot-action:
after: 30
command: "nova show goofy.demo"
post-stop-action:
command: "nova list"
pre-start and post-stop data are added into the output file.
periodic and single-shot are not because that would interfere with
the actual sampled data. Besides the intention is not to log statistics
but do things with the infrastructure such as server live migration.
TODO: add sections to the output file, something like pre, data & post
JIRA: YARDSTICK-46
Change-Id: Ia059813fb74733f86368aea9c7a20e5afb71d228
Signed-off-by: Hans Feldt <hans.feldt@ericsson.com>
|
|
A simple test case is added that will test writes to ephemeral storage.
See samples/fio.yaml
The Fio test type can also be used to test block storage, just modify
the "filename" argument.
JIRA: YARDSTICK-34
Change-Id: I1758d2999f8a5fdd44726e1dfc3e9769ea39dad6
Signed-off-by: houjingwen <houjingwen@huawei.com>
|
|
The "sequence" runner will use an input value of the test for every
run from a pre-defined list from the task file.
Example runner section of the task file:
type: Sequence
interval: 1
scenario_option_name: packetsize
sequence:
- 100
- 150
- 200
JIRA: YARDSTICK-47
Change-Id: I5bde9b78cb356499c338ef3da26ac1783670887d
Signed-off-by: Kristian Hunt <kristian.hunt@gmail.com>
|
|
JIRA: YARDSTICK-48
Step duration for Pktgen can be configured for each step.
Change-Id: Ida6e2f7d8a1c7d86fb2e65389b5b56d5958d3ad9
Signed-off-by: Kristian Hunt <kristian.hunt@gmail.com>
|
|
A simple test case is added that will ping an external server on
the internet. See samples/ping-ext-ip.yaml
Change-Id: I15eb3cb6ab9e5c1cf280f2aade2bf4c9646d6cd4
JIRA: -
Signed-off-by: Hans Feldt <hans.feldt@ericsson.com>
|