Age | Commit message (Collapse) | Author | Files | Lines |
|
We can read the openstack version from the env variables and then use
a different config. The reason we use this is because openstack API
changed from Pike to Queens and Pike is the official version for
OPNFV Fraser release
Change-Id: I8a94e71224d2ba0aa75a6438f3d7b03929acbbef
Signed-off-by: Manuel Buil <mbuil@suse.com>
(cherry picked from commit a459dee88ba00d6f8283232ce2159594ae23d22a)
|
|
|
|
Update the symmetric chain test case aligning it with updates available
from ODL Oxygen:
- ODL networking-sfc implementation now supports symmetric paths. It is
no longer needed to configure the ACL or chain directly through the
ODL rest api. It can now be fully relied on tacker for this.
- ODL networking-sfc implementation now supports different ingress and
egress ports for port pairs. The test case exercises this by setting
up a VNF with two distinct ports. Note that this requires an updated
VNF image that supports two nics and bundles a vxlan_tool capable of
using them.
Change-Id: Ie711abb93109943dcbf699ece7b2b570500a3711
Signed-off-by: Jaime Caamaño Ruiz <jcaamano@suse.com>
(cherry picked from commit 8e85daee26c523cd181f5884858646b7589c50af)
|
|
Starting Oxygen, ODL networking-sfc translator supports symmetric
paths and maps flow classifiers to ACLs with SFP redirect action
instead of RSP redirect action. A SFP may have two RSPs associated
if the path is symmetric, one otherwise. One of the two RSPs will
be flagged as the reverse path and the other is the forward path.
Traffic from the logical-source-port will be classified to the forward
path, while traffic from the logical-destination-port will be
classified to the reverse path.
This patch updates the ACL flow check logic to account for the above.
Change-Id: I2d8b9e6c2a1033d469668db29b9c18f525e89370
Signed-off-by: Jaime Caamaño Ruiz <jcaamano@suse.com>
(cherry picked from commit f66a8164cfb2ef172ee3ea274992b9c07306f08c)
|
|
There was a bug in releng and pylint was not ran, some issues got merged
in master because of that
Change-Id: Icf3fa882f4f2b0593fa0b85edce46e49dcc303c9
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
|
|
JIRA: SFC-122
Add support for Compass installer in SFC repo which will make
Compass SFC scenarios be verified by SFC testcase daily.
Change-Id: I8129ebe837d54cdcd163158990fccf40451b98ad
Signed-off-by: Yifei Xue <xueyifei@huawei.com>
|
|
Removing it was a mistake because without it is harder to debug (e.g. using
break points does not work when triggering tests from functest)
Change-Id: I3a958da0add0fb57946da333f75fb526c2fe6c33
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
It is no longer called via BashFeature.
Change-Id: I8ae6b24d5e02f269056c1d4615b436c1a2a3a903
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
Change-Id: Id524fbf1677844b5172f52c857cf8bed8d075ff3
Signed-off-by: Linda Wang <wangwulin@huawei.com>
|
|
SNAPs changed the API:
https://github.com/opnfv/snaps/commit/c711acf8ae3e4ad6f746500747857bcc9fd6f7be#diff-0b922a3e087121238d5a8b08ef5b894aR480
Now we require to pass keystone client to get the neutron port
Change-Id: I93fa4b127255c1a078aa3644773b7c8b886cea6d
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
env.get() ensures that default values are reused if unset.
It also avoids desynchronization between os.environ and CONST.
It follows the change "Get properly env vars or their default values"
[1] applied in Functest.
It also fixes minor issues on shebangs and takes several pylint rules
into account in this module:
- http://pylint-messages.wikidot.com/messages:w0702
- https://www.python.org/dev/peps/pep-0282/
[1] https://gerrit.opnfv.org/gerrit/#/c/52221/
Change-Id: Ia7ad168b2cbbade8ea3890fb95dc01bcdfc50468
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
The following tacker patch:
https://review.openstack.org/#/c/532203/
changed the syntax of the vnffgd templates
Change-Id: Ifbf788b02a95c8f45eb97230d3d2da117de158d6
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
This is the flow of the test:
1 - Create a vnffg in tacker (i.e. chain + classifier)
2 - Remove the vnffg
3 - Check that the RSPs are removed from operation DB in ODL and
the classifier rules are not present in the flows
4 - Create a new vnffg
5 - Test the new chain:
5.1 - Check HTTP traffic works
5.2 - Block HTTP traffic in VNF and check that it does not work anymore
Change-Id: Ia077e58c5cbce4258a2f8a36f6b961eb923ff9c5
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Change-Id: I11c37ed6898c03b2a54a0e3b11422c654023333f
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
When migrating to SNAPs a bug was introduced and could be seen in logs as:
Error when waiting for classification rules: 'module' object has no attribute
'get_compute_client'
The cause is that get_compute_client is not a global function anymore but a
method inside a class. Besides, due to SNAPs, we need to fetch the compute
name in another way
Change-Id: I32007c83894281ba9514a121b7939b5bfe52f4bc
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
|
|
Tacker is changing the config and adds cert_verify
Change-Id: Ia99bf2ad4ea0d6986ecd523bbe76e0e2b4a8a0ce
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
For Queens, tacker uses a different template syntax for the vnffgds
Change-Id: I3c3bd497f09fcf6af2adc0eaf0f046dad88d88f2
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Functest is deprecating its openstack library which we were using extensively
We need then to move to an alternative and the best appears to be SNAPs
Change-Id: Icaa0b9f1ec580545b9d0faa88be2080a310deaf7
Signed-off-by: Manuel Buil<mbuil@suse.com>
|
|
Due to this:
https://gerrit.opnfv.org/gerrit/#/c/49243/
We need to change the class
Change-Id: I17dc65fcfd3e7ce3f6ebe5ec5176e5d89ef2e0e2
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Change-Id: I1fca51b9b6afff5a596ceee2e1401127fe23231d
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Utils.py was getting messy. This patch divides it into three different files:
* test_utils.py
* odl_utils.py
* openstack_utils.py
The tacker library is integrated into openstack_utils.py
Change-Id: I310949d1cee49b6aa1c9b3396bf6d6ca458cbaac
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
It conforms with the change "Remove prepare_env" [1].
[1] https://gerrit.opnfv.org/gerrit/#/c/47267/
Change-Id: I9a80049e002b812d097067dc53798fa3bc532505
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
As agreed with the functest project, we keep the Dockerfile
and testscases.yaml in our repo
Change-Id: Idb25d4bfff7e36b42e4981a138775f90b0fbe72e
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Apex deployment shows a problem when creating the SFC config in ODL and
it is impossible to know what might be going on
Change-Id: If8d33d1c60daaf9cd9773a5cb9c7e47f64e1caa9
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
We moved to a new image for SFs but we have not moved the name of the image
in the vnfd
Change-Id: I8f658e1dfcf0de3117845f615eddb8fd5b1e11af
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
To create the four SFs we will use the sfc_nsh_euphrates image. This image is
like the sfc_nsh_danube but modifying the vxlan_tool which now has the
important fix:
https://git.opendaylight.org/gerrit/#/c/64116/
Change-Id: I396b5523b7243472b6d05b62d831837938067422
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
As agreed with the releng-xci guys, we should not use capital letters to refer
to openstack-ansible (osa)
Change-Id: I634747b07e72513f751cde3eb33f3b4bad5c176d
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
JIRA: SFC-116
It is confusing that our test-1 is in the second position of config.yaml
and that it uses test2-vnfd instead of test-vnfd
Change-Id: Iffd22f295630bce0ce854d4876c1478c56690cfa
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Test 2 is continuously failing because of tacker missing functionality
Change-Id: I785e79d78df4bfb9f48e20288e261efff741bcd4
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Change-Id: I64caba2eea3ff93d082c08915f82be7ef701a3f1
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
JIRA: SFC-114
Some cases show that it was not enough with the patch:
https://gerrit.opnfv.org/gerrit/#/c/42801/
We should explicitely disable security in the SF ports. That can be done
with a the antispoofing parameter in the vnfd
Change-Id: Ie20ab961633866f48a4abcc299a62e8c55280088
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Having the SFs port in security groups delays packets artifically in the tests
Besides, the netvirt guys suggested not to do it because it might be confusing
for ODL in some situations and it is not a 'suitable construct'
Tests without security groups show indeed a better behaviour
Change-Id: I33281215cd2e80804c8f09ca201c046f590249ab
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
JIRA: SFC-107
We are submitting the vnffgd with a variable for the neutron_port. Before
creating the vnffg, we must provide a file with the neutron_port value
Change-Id: I9994e2baa53f323df17434448d372598c0d94c72
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
JIRA: SFC-105
They have a typo which resulted in errors in Tacker when loading the template
Change-Id: I5723c756add7610377815180352140d58fd783a5
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
There are some checks in sfc_two... which do not exist in sfc_one. Those checks
should be done at the beginning of the test
Change-Id: I44c612c7dfe6bb4329bc49478f3342cda555229c
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Two things must be done:
1 - Be compatible with the new way of doing SFC in ODL Nitrogen
2 - Be installer independent
Change-Id: Ibbb3057db30a7c00e92f788bb03e770749e2b255
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Change-Id: I3a34daaca258cd2ce511aee51819cf438110e403
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
|
|
My VMs take around 80 seconds to be able to respond to pings. The current
timeout is in 50 seconds. I will increase it to 150 to have a buffer
Change-Id: Ie684e76799210e3d7239e35df2d4bdd7af507669
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
ODL requires that parameter to be present in order to classify
Change-Id: Idb8e0578a9c4c22cd35e4615755333b67d4fe658
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
|
|
|
|
In this patch [0] we adapt the Tacker library which is
located to the SFC repo to the new Tacker API
With the current patch we allign the testcases to the
aforementioned API
[0] https://gerrit.opnfv.org/gerrit/#/c/40839/
Change-Id: I07f8bcefbda7d058e9322465df38077ac0346d83
Signed-off-by: Dimitrios Markou <mardim@intracom-telecom.com>
|
|
Change-Id: I078e9f26d3040cf04120abd425f03cb4f2e8daef
Signed-off-by: Dimitrios Markou <mardim@intracom-telecom.com>
|
|
We are going to use ODL Nitrogen which does not require the NSH hack as
packets reaching the VNF will use ETH+NSH encapsulation
Change-Id: I5a55fde84e8f17860783ec9de8b99242fc38b9a9
Signed-off-by: Manuel Buil <mbuil@suse.com>
|
|
Change-Id: I710e6f5da92ddfddd9111c6c53274b92d76d7449
Signed-off-by: Dimitrios Markou <mardim@intracom-telecom.com>
|
|
Change-Id: I16a5e4522dde96de8d06e6cfb1e3a429b4f171de
Signed-off-by: Taseer Ahmed <taseer94@gmail.com>
|
|
Right now the testcases are dependent to only one installer.
So when we are trying to run the testcases to multiple installers
the tests are failing.
This patch make the testcases independent to the installer that they
run.
The testcases are running until the point which they hit the tacker
client commands.They are failing in the tacker client commands
because we need the new library for the upstream tacker API.
The testaces have already tested to two installers (Fuel,Apex-ha).
Also this patch contains the changes which are required for the
adaptation to the new functest.
For more information see the Jira tickets [0],[1],[2].
[0]: https://jira.opnfv.org/browse/SFC-100
[1]: https://jira.opnfv.org/browse/SFC-101
[2]: https://jira.opnfv.org/browse/SFC-102
Change-Id: Id1e5d5c94a65ab8bdea9584fa833bfa0cdec6632
Signed-off-by: Dimitrios Markou <mardim@intracom-telecom.com>
|