Sanjib Mohapatra [Tue, 12 Apr 2016 15:16:32 +0000 (20:46 +0530)]
Added test suites for cluster HA data recovery leader follower failover
Description: In a 3 node cluster initial inventory shard status is
verified and following tests are performed.
- Mininet switch is connected to a follower node and flow is added via
another follower node. Disconnect the switch and restarts the Leader.
Connect the switch again once Cluster is formed and verify flow is
installed in the switch again
- Disconnect the switch and restarts one of the follower node. Connect
the switch and verify the flow is installed in the switch again.
- Disconnect the switch and restarts the Cluster. Connect the switch
again when Cluster is formed and verify the flow is installed
successfully.
Change-Id: Ibd01664bc355a41bc95794930b1e40f3468ce858
Signed-off-by: Sanjib Mohapatra <sanjib.mohapatra@ericsson.com>
Jozef Behran [Thu, 7 Apr 2016 13:49:20 +0000 (15:49 +0200)]
Replaced NetconfViaRestconf.robot with TemplatedRequests.robot
NetconfViaRestconf.robot is not compatible with
ClusterManagement.robot and is obsoleted by
TemplatedRequests.robot. The incompatibility problem prevents
easy construction of test suites testing scenarios with node
outages so this outdated piece of code had to be replaced
with the newer (and compatible) piece of code. The change is
quite involved because TemplatedRequests.robot has quite
different API.
Change-Id: Iaa791eab06d1eaaf4d1904f248c57f67a30afc61
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Tue, 12 Apr 2016 08:40:39 +0000 (10:40 +0200)]
Tidied updated test suites
Change-Id: I03ce220e896169f145674943e2e4f6fe430bf6da
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Samuel Kontriš [Wed, 6 Apr 2016 13:45:20 +0000 (15:45 +0200)]
Topoprocessing - test correction according to patch 36891
Patch 36891 is bugfix for bug 5458
Change-Id: Id859b7afa1ee0c4cf3e502524df6d97064ecd1d1
Signed-off-by: Samuel Kontriš <samuel.kontris@pantheon.tech>
Ruchita [Thu, 31 Mar 2016 09:22:29 +0000 (14:52 +0530)]
Add csit test cases for centinel features.
Change-Id: I65d4ea2efbeea3ba1aaa9d7044b4b1005675d5bc
Signed-off-by: Ruchita <ruchita.khanna@tcs.com>
Vratko Polak [Fri, 8 Apr 2016 12:20:44 +0000 (14:20 +0200)]
Bug 5560: Fix bits ordering issue in BGP flowspec tests
This includes change to norm.json.py Library to optionally
perform sorting withing string values if key is present in given list.
Also, ordering and syntax inside Verify Flowspec Data is improved.
Change-Id: I12d2493233d557772f8beec6e11743a4d6941c97
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Jamo Luhrsen [Fri, 8 Apr 2016 04:29:14 +0000 (21:29 -0700)]
enable auth in rest calls, and fix one error message
Change-Id: Iddd0dfb7ec44a89c00f3f481a5576459b113077e
Signed-off-by: Jamo Luhrsen <jluhrsen@redhat.com>
Jozef Behran [Thu, 7 Apr 2016 11:22:59 +0000 (13:22 +0200)]
Add ability to use JAVA_HOME to find java
If no JDK_VERSION is specified and invoking "java" on the
commandline does not work at all ("java: command not found"),
the code uses the value of the JAVA_HOME as the root of the
JDK to be used as a last resort.
Change-Id: Ieb647003dc7fe4790b747191371ae240fec40d1b
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Tue, 5 Apr 2016 12:58:54 +0000 (14:58 +0200)]
Fix malformed BASH command in ComposeDilemmaFilePath
A misplaced ampersand caused the command to always return 0
(success), even when the path queried did not exists and the
command was thus supposed to return nonzero (a failure).
Change-Id: If59501fa7d79fc60f0f63fde02095a1c87b19e59
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jamo Luhrsen [Wed, 6 Apr 2016 21:03:34 +0000 (14:03 -0700)]
prepend ovsdb to connection-info for NB connection request
for the case when hwvtep features are loaded along side of
ovsdb, the connection-info data is taken from the same base
model, so there is a conflict resulting in an ERROR when
trying to initiate a connection with just "connection-info".
we need to make it more specific, like "ovsdb:connection-info"
Change-Id: I5729a9c1d91507dd3dd6e2cfb8591e916aa6fa88
Signed-off-by: Jamo Luhrsen <jluhrsen@redhat.com>
Vratko Polak [Wed, 6 Apr 2016 14:24:20 +0000 (16:24 +0200)]
Fix typo in throughpcep suite
Change-Id: I76dbfb36a0b69ec0f7a8c50a5ffb9c0b942776c3
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Daniel Farrell [Mon, 4 Apr 2016 18:28:06 +0000 (14:28 -0400)]
Ignore all vim tmp file extensions
After .swp, vim roles over to other .sw* files. Ignore all of them.
This is the same logic we use in int/pack/.gitignore.
Change-Id: Id9cac1cc65aa919993336b079c068e9ef7bd67ee
Signed-off-by: Daniel Farrell <dfarrell@redhat.com>
Luis Gomez [Fri, 1 Apr 2016 01:40:12 +0000 (18:40 -0700)]
Comment last OF cluster test to better observe bug 5641 in karaf log
In any case current HA suite covers the remove Sanity check.
Change-Id: Ica757008c42589b0af3a6bab8992b834ff931e9c
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Luis Gomez [Thu, 31 Mar 2016 22:19:56 +0000 (15:19 -0700)]
Adding more time to stabilize CPqD tests
If this does not help we will consider to remove all CPqD tests from CI.
Change-Id: Ic03ebf3c65abe588020f33444bcb5fef589305ab
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Miguel Angel Munoz Gonzalez [Thu, 31 Mar 2016 12:49:43 +0000 (14:49 +0200)]
Fix SFC robot test cases in CSIT
This commit is composed of two different changes:
1. Replace ODL_VERSION variable with ODL_STREAM in SFC test cases
ODL_VERSION variable in robot test cases is no longer used and instead
ODL_STREAM is the new name, with the same content.
2. Fix SFC_FUNCTIONS_FILE variable definition in SFC Robot test cases
Definition of variables in robot test cases need to keep separation of
four spaces between the var name and the value
Tested in Jenkins Sandbox: https://jenkins.opendaylight.org/sandbox/job/sfc-csit-3node-clustering-all-beryllium/5/
Change-Id: I37b326d92f534411d45e92f88b2e0f120e6ad506
Signed-off-by: Miguel Angel Munoz Gonzalez <miguel.angel.munoz.gonzalez@ericsson.com>
Sanjib Mohapatra [Mon, 4 Apr 2016 10:06:21 +0000 (15:36 +0530)]
Added test suites for cluster HA data recovery leader follower failover
Description: In a 3 node cluster initial inventory shard status is
verified and following tests are performed.
- Mininet switch is connected to a follower node and flow is added via
another follower node. Disconnect the switch and restarts the Leader.
Connect the switch again once Cluster is formed and verify flow is
installed in the switch again
- Disconnect the switch and restarts one of the follower node. Connect
the switch and verify the flow is installed in the switch again.
- Disconnect the switch and restarts the Cluster. Connect the switch
again when Cluster is formed and verify the flow is installed
successfully.
Change-Id: I1cc767712bac3694ba4bf9b052765903bce28ae8
Signed-off-by: Sanjib Mohapatra <sanjib.mohapatra@ericsson.com>
Daniel Farrell [Tue, 8 Mar 2016 00:21:28 +0000 (19:21 -0500)]
Add requirements.txt for cluster-deployer tool
Using `pip install -r requirements.txt` against this file, ideally in a
virtual environment, will install the dependencies of `deploy.py` and
its helper scripts.
Change-Id: Ib70f617d7ecf3fa14afba22adfe39c2363a36d1c
Signed-off-by: Daniel Farrell <dfarrell@redhat.com>
Daniel Farrell [Mon, 4 Apr 2016 19:17:26 +0000 (15:17 -0400)]
Fix cluster-deployer script execution in venv
The hard-coded path to the python executable in deploy.py and restart.py
broke when executing them like scripts in virtual environments.
```
$ ./deploy.py
zsh: ./deploy.py: bad interpreter: /usr/local/bin/python: no such
file or directory
```
Moves to standard `/usr/bin/env python` method.
Removes extra comment character from just before restart.py shebang.
Marks restart.py as executable, since it's designed to be run like a
script.
Change-Id: I945e5704327fb8a2c6387a919054560bd23055d1
Signed-off-by: Daniel Farrell <dfarrell@redhat.com>
Daniel Farrell [Tue, 8 Mar 2016 02:45:01 +0000 (21:45 -0500)]
Add requirements.txt for cluster-monitor tool
Using `pip install -r requirements.txt` against this file, ideally in a
virtual environment, will install the dependencies of `monitor.py` and
its helper scripts.
Change-Id: I5769fbcbf9db4d9caeed56707b0c968fa57b38be
Signed-off-by: Daniel Farrell <dfarrell@redhat.com>
Jozef Behran [Mon, 4 Apr 2016 13:04:16 +0000 (15:04 +0200)]
Fix scaling tests failing due to a name clash
The tests get wrong timeout due to a clash with
NetconfKeywords which uses the same name for a variable with
testtool device boot timeout. Rename that NetconfKeywords
variable to better indicate its purpose (and thus to avoid
the name clash).
Change-Id: I255308d5b2b754f632985166de7b25a8dd43b194
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Mon, 4 Apr 2016 11:30:27 +0000 (13:30 +0200)]
Tell CURL to follow redirects
After moving Nexus to the private cloud the tests have been
broken because CURL does not follow redirects by default.
Change-Id: I82ec9bd74527ee5bfd15879a7730b55cb88522ec
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Luis Gomez [Thu, 31 Mar 2016 03:35:40 +0000 (20:35 -0700)]
Add WUKS and stats collection time in OpenFlow perf test.
Change-Id: Ied3f5cffac7686c891fe0991aafa299eff3920e9
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Mohamed El-Serngawy [Fri, 1 Apr 2016 16:15:19 +0000 (12:15 -0400)]
Adding -o option to unzip odl steps in the deploy.py script,
without -o option the script keep waiting for interrupt and
didn't proceed.
Change-Id: I9a45cb1777148fa2b6e963d6697fa76aff836a07
Signed-off-by: Mohamed El-Serngawy <melserngawy@inocybe.com>
hari.pr [Tue, 8 Mar 2016 11:41:19 +0000 (17:11 +0530)]
Script to check cluster status and entity owner
* Required jolokio bundle installed in odl
* Run the script "python cluster_check.py"
* Provide the ip address of the controller to get the details
Change-Id: I94c6bbe5c2ab53b3562419b6769018e77f50210c
Signed-off-by: hari.pr <hari.pr@hcl.com>
Sanjib Mohapatra [Tue, 8 Mar 2016 11:37:26 +0000 (17:07 +0530)]
Adds test suits for Cluster HA Leader follower failover
Description: Flow add modify and delete at leader or follower also test same
after leader follower or cluster restart
Initial inventory shard status test among cluster nodes. Switch connected to
leader or follower node and verify flow add, modify, delete in leader or
follower nodes. Similarly verify RPC flow add, modify, delete in leader as
well as follower nodes. Reiterate the same functional test after leader
restart, follower restart and cluster restart.
Following changes are incorporated.
- Wrote good and meaningfull commit message
- Removed "Kill Specific Controller" and "Start Specific Controller" instead
used alredy existing "Kill One or More Controllers" and "Start One Or More
Controllers" in ClusterKeywords
- Used tidy tools for the checked in files
- Revert back the timer to original 15secs in ClusterOpenFlow
for flow add modify
- Removed ${KARAF_HOME} ${WORKSPACE}/${BUNDLEFOLDER} from test suite file
- Modified keyword "Start Multiple Controllers" so that first all nodes
restarts and then "Wait For Controller Sync" performed in all controller
nodes
Change-Id: I62d9f6c487d591b21b1fccdda9a5618ed3df5295
Signed-off-by: Sanjib Mohapatra <sanjib.mohapatra@ericsson.com>
Peter Gubka [Thu, 31 Mar 2016 15:54:37 +0000 (17:54 +0200)]
changing regexp for simple substring search
the regexp did not work
Change-Id: If5dfd3cd1c77d034f887a8dd024f97fc32fec88f
Signed-off-by: Peter Gubka <pgubka@cisco.com>
manimuthazhagan [Fri, 15 Jan 2016 11:42:20 +0000 (17:12 +0530)]
Fix Netvirt Tests by splitting the single node and three node tests
* Created a new TestPlan for 3-node tests
* Created a set of tests for 3-node
* Deleted a Older Ovsdb_Net-virt Test cases
* Removed the essentially noop suite (Ovsdb_Netvirt_flow) as this patch
enables it in cluster mode
Change-Id: I3fd9e2574c8960f3ba3690dceac4d9f122d93494
Signed-off-by: manimuthazhagan <manimuthazhagan.r@hcl.com>
Signed-off-by: Jamo Luhrsen <jluhrsen@redhat.com>
Samuel Kontriš [Thu, 24 Mar 2016 17:12:48 +0000 (18:12 +0100)]
Topoprocessing - corrected tests for filtration and unification
of termination points on inventory model
According to change https://git.opendaylight.org/gerrit/#/c/36507/3
Change-Id: I7c2bfcf56f0411922eac77e59c4c78303918fcf8
Signed-off-by: Samuel Kontriš <samuel.kontris@pantheon.tech>
Samuel Kontriš [Wed, 23 Mar 2016 15:47:33 +0000 (16:47 +0100)]
Topoprocessing - corrected tests for filtration on links
Change-Id: I7bd7235c256ce7965003d30b8ab49be5dea6c262
Signed-off-by: Samuel Kontriš <samuel.kontris@pantheon.tech>
Karthik.S [Thu, 10 Mar 2016 18:21:01 +0000 (23:51 +0530)]
Modify CSIT test files and library file of VTN Manager.
* Fixing invalid input values(inet4src, inet4dst, icmp code and icmp type) in flowfilter robot test files.
* Modified Dumpflow command in VtnMaKeywords.robot.
* Due to Bug 5587 added Tags exclude for ICMP test cases in flowfilter robot files for both OF10 and OF13.
Once bug is fixed and resolved, will remove Tags exclude and enable ICMP test cases in flowfilter robot files.
Change-Id: I7dec55e97affc0ccfb56c2104faa34b541646cee
Signed-off-by: Karthik.S <siva-karthik@hcl.com>
Martin Mihálek [Mon, 21 Mar 2016 15:08:31 +0000 (16:08 +0100)]
Boron Update of SXP CSIT.
Change-Id: I900b2aad077133d0284200228de19711ba985dcf
Signed-off-by: Martin Mihálek <mamihale@cisco.com>
Luis Gomez [Mon, 28 Mar 2016 19:59:44 +0000 (12:59 -0700)]
Consolidate Entity Owner keyword in ClusterKeywords.robot library
Change-Id: Ie435bda9c469d9bed5aedd50d978cdb85d5a5ee7
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Luis Gomez [Wed, 30 Mar 2016 03:02:37 +0000 (20:02 -0700)]
Change flow count criteria packet-count to priority
There is an extra packet-count for table aggr stats in Li branch.
Change-Id: I7269739d759fef2087174705460f1c3ead327909
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Vratko Polak [Wed, 23 Mar 2016 15:48:39 +0000 (16:48 +0100)]
Remove BUNDLEURL variables
Csit jobs already provide -v option with the correct value,
based on .zip under test.
Without this change, people lauching robot scripts manually
would get unexpected value (Lithium pre-release snapshot version)
instead of error due to undefined variable
(acting as a notification that they need to specify
the correct value when starting pybot).
Change-Id: Idaad1a46474cc9ea917c4f78845231141e0b5dfe
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Luis Gomez [Mon, 28 Mar 2016 18:34:40 +0000 (11:34 -0700)]
Remove AD-SAL tests from int/test repository
Change-Id: Idbaa886c2e2afa96d5f96a4b4c4edec2723981b2
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Jamo Luhrsen [Tue, 29 Mar 2016 05:11:18 +0000 (22:11 -0700)]
removing AD-SAL location from testplans
Change-Id: Ib8e667702a71ff17ab3b6e8171e28c789ee0a41e
Signed-off-by: Jamo Luhrsen <jluhrsen@redhat.com>
Praveen Mala [Sat, 16 Jan 2016 00:30:54 +0000 (19:30 -0500)]
Adding port delete functionality to ovsdbconfigblaster.py
Change-Id: I93d552fc774390846913fea74b2500da89899d79
Signed-off-by: Praveen Mala <praveen.mala@intel.com>
gvrangan [Tue, 22 Mar 2016 01:34:39 +0000 (01:34 +0000)]
Fix for VTN Manager Li job Failure
Removed the method to add flows for table miss,
it looks like Jenkins is using the OVS 2.0.1 for
Li jobs and this is not necessary.Retained the keyword
so that it can be used in the future
Change-Id: I7f8a6c03a29737dadbaae659fd7fd6b3fc9fe034
Signed-off-by: gvrangan <venkatrangang@hcl.com>
Jozef Behran [Wed, 23 Mar 2016 17:57:18 +0000 (18:57 +0100)]
Add code to kill restperfclient if hung.
Change-Id: Ifc077926b1aa1a1eb1c7d1b38f95a14a6a3cc977
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Thu, 24 Mar 2016 09:58:46 +0000 (10:58 +0100)]
Refactor performance suites to use keyword timeouts
This removes the ugly hack with ${XXX_TIMEOUT_FOR_TEST_CASE}
global variables and test timeouts for a more elegant
solution that uses a keyword timeout instead. Requires Robot
Framework 3.0.
Change-Id: If5876cce5edccfc693a06fff5a9c8c892b2da0f2
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Thu, 24 Mar 2016 10:07:40 +0000 (11:07 +0100)]
Remove dead code that somehow survived code cleanup
Change-Id: I4fef816734be3eeb0b6ff277d7a8b458c9da359d
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Thu, 24 Mar 2016 09:54:58 +0000 (10:54 +0100)]
Add missing reference to SetupUtils
Change-Id: Ib308ea9439a77f0ac88e7ffb2998afd23a61ced1
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Wed, 23 Mar 2016 17:52:33 +0000 (18:52 +0100)]
Tidied updated resource
Change-Id: If74ebb7276a3dbda64dd2e50e488ff9aca55e213
Signed-off-by: Jozef Behran <jbehran@cisco.com>
priya.ramasubbu [Mon, 29 Feb 2016 13:13:40 +0000 (18:43 +0530)]
Single Node Odl Tempest Job Tests
* Create two networks, two subnets and router
* Verify dhcp, gateway and instance ips in the dump flow
Change-Id: I714ccaf83d52aea7a8494a0c10efc5c3a4e83a48
Signed-off-by: priya.ramasubbu <priya.ramasubbu@hcl.com>
Jozef Behran [Tue, 22 Mar 2016 09:46:23 +0000 (10:46 +0100)]
Configure timeout in RestPerfClient
Add --timeout argument setting the timeout for the
RestPerfClient threads issuing requests to the timeouts
calculated in the suites instead of the default 5 minutes.
Do that change for all three RestPerfClient performance
tests.
Change-Id: I7497eb4284b3dde5909d1fab68783d03ce7a5e1a
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Wed, 23 Mar 2016 07:52:25 +0000 (08:52 +0100)]
Remove restperfclient invocation code triplication
Move the triplicated restperfclient invocation code into its
own bunch of keywords and hide some implementation details
from the test suite files. This makes the test suite files
easier to read and as a result it is also easier to add
additional code into the restperfclient invocation process.
Change-Id: I7a64b15e72c5fcf7766a768b44d6ebc3dd5434b2
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Wed, 23 Mar 2016 07:38:01 +0000 (08:38 +0100)]
Move the timeout tests in mdsal.robot to their proper place
The timeout tests are conceptually "failed requests tests"
so they don't belong to a test case whose name says "invoke
restperfclient" when there is another whose name says "check
for request failures". Move these checks to the proper test
case.
Change-Id: I7ab625df6c29fd6fb91b7174715476ed77eda07f
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jamo Luhrsen [Tue, 22 Mar 2016 18:03:02 +0000 (11:03 -0700)]
cleaning up system variables
- trying to change all MININET to TOOLS_SYSTEM as well as CONTROLLER
to ODL_SYSTEM
- also found a few small refactors to make
* MODULES_API instead of REST_CONTEXT
* moved DIDM variables to subdir for clarity
Change-Id: Ifa50e14a6f9479c55dfd0a748fb8936e1a6af0a6
Signed-off-by: Jamo Luhrsen <jluhrsen@redhat.com>
Jozef Behran [Tue, 22 Mar 2016 09:39:23 +0000 (10:39 +0100)]
Add test for RestPerfClient timeouts
When RestPerfClient times out, it means a poor Netconf
Connector performance bug was caught. In the MDSAL Northbound
case it is reported as bug 5581, no such bugs were seen in
the other two cases so none are reported. However after the
RestPerfClient bug that causes it to hang in this situation
(see bug 5413) is fixed, the suite in itself tends to fail to
take a notice of such problems so add this specific check for
timed out threads that points the poor Netconf performance
bugs out.
Change-Id: Ib75707f517ee34f40b6c9961f90309c26870bafb
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Tue, 22 Mar 2016 09:55:39 +0000 (10:55 +0100)]
Add missing timeout to the test tool device performance test
The RestPerfClient on testtool device test suite was missing
a timeout setting. This would lead to suite hangs when this
scenario would trigger a "poor Netconf connector performance"
bug (see bug 5581 for an example) or spurious successes.
Change-Id: Ibf1c1843a86de12e11a4012c4209a36c0f0997e6
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Tue, 22 Mar 2016 15:26:20 +0000 (16:26 +0100)]
Tidy updated test suites
Change-Id: I8005dc69c8d07380bcdfefb832ebffb7c4d698e8
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jamo Luhrsen [Mon, 21 Mar 2016 22:31:45 +0000 (15:31 -0700)]
Giving the right library for Report_Failure_Due_To_Bug_Found keyword
Change-Id: I9d72f9cbe0b0e89b1f188e2034e94be171f17b11
Signed-off-by: Jamo Luhrsen <jluhrsen@redhat.com>
Jamo Luhrsen [Mon, 21 Mar 2016 22:11:56 +0000 (15:11 -0700)]
Updating ODL versions for consumption by networking-odl project
Change-Id: I0141da5aea7daac874a1f6efc4a60c9fa2e11790
Signed-off-by: Jamo Luhrsen <jluhrsen@redhat.com>
Jozef Behran [Mon, 21 Mar 2016 11:22:46 +0000 (12:22 +0100)]
Add PIP installation command into notifications test suite
Some tools system images (Ubuntu in particular) don't have
PIP installed and this causes the notifications test suite to
fail. Fix that by adding the code to install PIP before using
it.
Change-Id: Ib871cf554fc9c8e9456e16882d799a532fbf2c8d
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Luis Gomez [Sat, 19 Mar 2016 00:58:23 +0000 (17:58 -0700)]
Add more robust wait for flow stats
Change-Id: Iae0c0d26d6f1b8ee54d67f236d51c811567cb5a3
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Karthik.S [Mon, 14 Mar 2016 11:09:31 +0000 (16:39 +0530)]
Modify VTN Manager CSIT for Stable Lithium
* Add Flow to OpenFlow Switch to forward packet in case of Table Miss
* Fix the wrong Keyword that was causing the entire job to fail
Change-Id: I32a66208acb88b66d178ba87d63135922b01840c
Signed-off-by: Karthik.S <siva-karthik@hcl.com>
gvrangan [Fri, 18 Mar 2016 22:18:58 +0000 (22:18 +0000)]
Fix VTN Coordinator CSIT Failures
* Removed Unwanted Call to Stop Mininet
* The Mininet wqs stopped by an earlier case
Change-Id: Ia106b348cdc4d7db52befd59b6ceb96808f7c369
Signed-off-by: gvrangan <venkatrangang@hcl.com>
Peter Gubka [Fri, 18 Mar 2016 14:37:26 +0000 (15:37 +0100)]
renaming bgp userfeatures testplans to fit 35774
Change-Id: Ifca4fb70720a7dadd559db46ee23e7e1c1b096d0
Signed-off-by: Peter Gubka <pgubka@cisco.com>
Vratko Polak [Thu, 3 Mar 2016 17:46:48 +0000 (18:46 +0100)]
Rework of controller car/people cluster suite
- Deleted the previous car/people cluster suites
+ New Resources: CarPeople, ClusterManagement
+ Updated/fixed: TemplatedRequests.robot, norm_json.py
+ New suite names, without number prefixes
+ Full-fields data files for car/people TemplatedRequests
- Simplistic iterated values for the fields
+ controller-clustering testplan updated
Change-Id: I56d0743129d6e8108b66e9b48c135e4c3bc13d5c
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
vinoth [Mon, 29 Feb 2016 22:27:53 +0000 (03:57 +0530)]
Bug 5423: Fix for ERROR messages when intents were deleted.
* Previously the warn message "Invalid Address" is written in source code as log.error,
due to this warn message was displaying in error message.
* Currently the warn message handled properly in the source code as log.warn
link: https://git.opendaylight.org/gerrit/#/c/35273/
other changes:
* Fixed one of additional test case failed on to check drop the
packet when intent is delete.
Change-Id: I608ba47c8b64639c09a22056a710be194ca6bc50
Signed-off-by: vinoth <m.vinoth@hcl.com>
Karthik.S [Tue, 15 Mar 2016 11:24:17 +0000 (16:54 +0530)]
Enable loglevel as Trace for VTN Manager
Change-Id: Ie118580c68cab3e11bd6203b39eeb0a35fcc4257
Signed-off-by: Karthik.S <siva-karthik@hcl.com>
Peter Gubka [Thu, 11 Feb 2016 10:42:25 +0000 (11:42 +0100)]
flowspec tests
it has two test cases
-various flow details
-redirect option
Change-Id: I440ba40669711092917b9250e6546c3f8c891ce8
Signed-off-by: Peter Gubka <pgubka@cisco.com>
Tomas Cechvala [Tue, 15 Mar 2016 15:04:03 +0000 (16:04 +0100)]
Creating testplan for Openstack job
Change-Id: Ic356d9aafb2a4d16912c7e10814e9e43bcd6a63a
Signed-off-by: Tomas Cechvala <tcechval@cisco.com>
Luis Gomez [Wed, 16 Mar 2016 22:11:06 +0000 (15:11 -0700)]
Replace sleep by WUKS in Flow Stats suite
Change-Id: I99e4849f965293d938751f7ce0bbb99be735d60c
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Vratko Polak [Tue, 15 Mar 2016 14:39:43 +0000 (15:39 +0100)]
Include needed Assure_* keywords in BGP tests
Ubuntu tools system seem to be missing ipaddr installation.
This could be fixed on releng/builder side,
but this makes it easier to run on onther environments.
Also, small improvements in Keyword ordering in BGP suites.
Change-Id: Ic863a083d0306ce5c9a7cc8b1f89c061a1e92065
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Radovan Sajben [Thu, 18 Feb 2016 11:59:40 +0000 (12:59 +0100)]
Carrying LSP State Information in BGP
- basic functional test
Change-Id: I1866d7d96dc4265c595d709714bc381a2b5ff61b
Signed-off-by: Peter Gubka <pgubka@cisco.com>
Jozef Behran [Mon, 14 Mar 2016 11:30:30 +0000 (12:30 +0100)]
Fix documentation of Utils.Set_User_Configurable_Variable_Default
The documentation was updated to stress that the name of the
variable to be set must be specified without the ${} syntactic
sugar. Also the TODO was updated to stress we don't have yet
examples of usage where doing the TODO would bring a benefit.
Change-Id: I614769c9c2565c0463633d56b306510c829e3965
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Lorand Jakab [Tue, 15 Mar 2016 18:35:59 +0000 (20:35 +0200)]
Add auth key for 0.0.0.0/0 in LISP perf tests
We recently discovered a security issue in lispflowmapping and an
explicitly configured authentication key is needed since the fix was
committed to allow southbound Map-Register messages to write into the
mapping database. This patch adds such a key for the Map-Register tests
to produce usable results (notifies/s graph has value 0).
Change-Id: I14bac6c92f040abce71f40a24c531b9ca8e312b3
Signed-off-by: Lorand Jakab <lojakab@cisco.com>
Luis Gomez [Wed, 16 Mar 2016 04:36:24 +0000 (21:36 -0700)]
Add WUKS to stabilize l2switch suite
Change-Id: I426ea63f67d690a9782df1b9c0e4dbb08c3af5e2
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Jozef Behran [Fri, 19 Feb 2016 13:25:38 +0000 (14:25 +0100)]
Performance test suite for MDSAL northbound
This suite uses RestPerfClient on MDSAL northbound, via
Restconf. Similar to "performance.robot" but no device is
involved and thus no testtool deployment is necessary,
the only needed tool is restperfclient.
Change-Id: I117e790cd5434933d105ccbc29c6bcf7a9fca69e
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Vratko Polak [Tue, 15 Mar 2016 11:13:03 +0000 (12:13 +0100)]
NexusKeywords: Make Compose_*_Java_Command also work on Ubuntu
Active SSH connection is now required,
but all suites had that satisfied already.
Change-Id: Ib0222ce50358e122c288bc4d74ac2026059bf714
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Jozef Behran [Thu, 10 Mar 2016 17:10:34 +0000 (18:10 +0100)]
Tidy new and updated test suites
Change-Id: I1290e2eff4977de37bd8375c380ca7cd1161c190
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Samuel Kontriš [Wed, 9 Mar 2016 15:15:29 +0000 (16:15 +0100)]
Topoprocessing Basic tests - Link computation
Change-Id: I660acae83a253fd18aad6731f9c18e18db2816e6
Signed-off-by: Samuel Kontriš <samuel.kontris@pantheon.tech>
Lorand Jakab [Tue, 8 Mar 2016 23:10:58 +0000 (15:10 -0800)]
Add Map-Register LISP SB performance test
Change-Id: I1c0a4a599733d3e3b5c24756fcfcf116e8db5dcd
Signed-off-by: Lorand Jakab <lojakab@cisco.com>
Jozef Behran [Fri, 13 Nov 2015 09:30:30 +0000 (10:30 +0100)]
Cycling device and checking that the data survived.
After the device data modification step the test suite
deconfigures the device, configures it back and checks that
the data is still there as they were when it was last
modified. Then the suite tries to modify the data again
just to make sure the deconfigure/configure cycle did not
mysteriously make them read-only.
Change-Id: I403357d85b80f28060775e57bb6192de90a6114e
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Fri, 4 Mar 2016 10:46:12 +0000 (11:46 +0100)]
Clear known bug ID in SetupUtils.Teardown_Test_Show_Bugs_...
When a known bug ID is set to e.g. 1234 and the test steps
after that will catch the bug 1234, the bug ID 1234 is
reported (correctly). But then the rest of the test case is
skipped and the next one is started. This means that the
next test case will have the bug ID set to 1234 and if
it fails, it will also report it found bug 1234 while in
reality it was supposed to report a bug search link instead.
Fix the problem by clearing the bug ID after reporting the
bug.
Change-Id: I44d9c17bd9e22dc4792529dea0425a7f6d06940e
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Samuel Kontriš [Fri, 26 Feb 2016 12:30:26 +0000 (13:30 +0100)]
Topoprocessing - tests correction
- reworked aggregation with filtration tests
- reworked Filtration termination points tests
- removed unnecessary files
- corrected some other tests
Change-Id: I8fc28154433517f22965d5eb93a52f218f2401ed
Signed-off-by: Samuel Kontriš <samuel.kontris@pantheon.tech>
Karthik.S [Thu, 10 Mar 2016 19:13:04 +0000 (00:43 +0530)]
Fix for VTN Manager CSIT test failures for Beryllium.
* Removed "Stop Suite" Keyword from Stop SuiteVtnMa function. Since, "Stop Miniet" Keyword used at the end of the VTN suite
which is used again in suite "TearDown function" which makes Test suite fails.
Change-Id: I1e5cf84ba7a76bb19bf4d7dc8f11bcdd12a58d56
Signed-off-by: Karthik.S <siva-karthik@hcl.com>
Jozef Behran [Thu, 10 Mar 2016 16:17:41 +0000 (17:17 +0100)]
Fix more error message problems
It turns out there is much more code duplication regarding
error message reporting. Fix that so the error message
strings appear in the code only once and are reused for
both, the exception and the log file. As a result the new
code is more readable and less confusing.
Change-Id: Ib72b7de11648b718e2db41ee1d576ad1993cbc10
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Thu, 10 Mar 2016 15:56:10 +0000 (16:56 +0100)]
Do not emit raw binary data after unconfirmed open
When the OPEN message generated by the player is rejected,
the original code emitted raw binary data in the resulting
exception. Change that so it emits the hexlified version of
the raw binary data. The fix was done by removing the
duplicated code that generates the two error messages (one
for the log file, the other for the exception), taking the
correct code only.
Change-Id: If1ac202ba0d3c8720c8c442da1d533143dcddb7c
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Samuel Kontriš [Thu, 25 Feb 2016 13:20:27 +0000 (14:20 +0100)]
Topoprocessing - bugfix for bug5156
Change-Id: Iaf9e7aadbc05bc49400286fe5c37f814ecc7650c
Signed-off-by: Samuel Kontriš <samuel.kontris@pantheon.tech>
Vratko Polak [Wed, 9 Mar 2016 13:02:21 +0000 (14:02 +0100)]
Fix logic in updater.py with respect to Counter class
Previously, it worked for Counter.py but not for collections.
Change-Id: I0fff144b0c261b47ce40a03766281f9df642f277
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Luis Gomez [Thu, 10 Mar 2016 05:28:10 +0000 (21:28 -0800)]
Add WUKS after conneting OVS instance to controller
Change-Id: Ie2fdeee11b851fa2237f6c752ead0d6aa4266484
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Vratko Polak [Wed, 9 Mar 2016 14:38:38 +0000 (15:38 +0100)]
Throughpcep: Download pcc-mock log to current working directory
Change-Id: I9d5657fab8b831e9e6bafd51775d3950a4801b34
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Vratko Polak [Wed, 9 Mar 2016 10:51:07 +0000 (11:51 +0100)]
Fix typo in throughpcep suite.
Change-Id: Ia5152cd9c5c6f04d4c2c9b95a35a8991c9a8a2ad
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Peter Gubka [Tue, 8 Mar 2016 14:44:01 +0000 (15:44 +0100)]
adding onos benchmarking scripts
Change-Id: I505ba2f1085392fa5031816c6fc1c7e2f50a0fd2
Signed-off-by: Peter Gubka <pgubka@cisco.com>
Jamo Luhrsen [Tue, 8 Mar 2016 00:31:41 +0000 (16:31 -0800)]
adding more configuration around shard names and restconf port
- added ability to provide specific port numbers for each controllers
restconf endpoint
- added ability to selectively ignore some shards. specifically, I
didn't want to look at the toaster shard when giving a demo
Change-Id: Ia7186193d186e34ba7963741ac54902d9619d34f
Signed-off-by: Jamo Luhrsen <jluhrsen@redhat.com>
Jan Medved [Mon, 7 Mar 2016 00:13:27 +0000 (16:13 -0800)]
Fixed flow_config_blaster Floodlight edition
Fixed measurements of time for stats collection in flow_add_delete_test
Fixed PIP-8 violations (at least the fixable ones)
Change-Id: I01a361a7245fe479fbb65eea95bea93808d787a8
Signed-off-by: Jan Medved <jmedved@cisco.com>
Luis Gomez [Sun, 6 Mar 2016 21:19:53 +0000 (13:19 -0800)]
Remove exclude cleaning from OVSDB cluster test case
Change-Id: I5aaa4551d8a470aff115a0dd42f1504deffe1bd6
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Jozef Behran [Thu, 3 Mar 2016 14:30:27 +0000 (15:30 +0100)]
Make pep8 more picky
Several guidelines were disabled and as a result I was
slipping into my old coding style, incompatible with the
Python sources here. Enabling them turned out to uncover I
was not the only one slipping. As a matter of fact all
checks are enabled now except E133 which is not compatible
with E123.
Change-Id: I754c740f49a213360df5b187897e776266e6653f
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Luis Gomez [Mon, 29 Feb 2016 05:45:30 +0000 (21:45 -0800)]
Fix and clean OVSDB cluster suite
- Simplify OVSDB cluster keywords.
- Modify Entity Owner KW to support uuid.
- Restore OVSDB clean environment.
Change-Id: I2c3034c5d09a4fce4e0b96f40bced30fff3204dd
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
premnath [Mon, 29 Feb 2016 11:40:05 +0000 (17:10 +0530)]
Modified HBase scripts to run in Lithium Branch.
Added Testplan to source.
Change-Id: I7fe969340d95c4bdf15f2ea836498096f3de58aa
Signed-off-by: premnath <prem.annauniv@gmail.com>
Jozef Behran [Fri, 26 Feb 2016 11:00:49 +0000 (12:00 +0100)]
Add keyword needed by guidelines
The new variable naming guidelines [1] specify the use of a
new keyword named Utils.Set_User_Configurable_Variable_Default
when the default value for an user configurable variable
needs to be computed at the suite setup, either using
BuiltIn.Evaluate or even invoking other, more complex
keywords (that look around in the environment or do other
stuff like that).
[1] https://wiki.opendaylight.org/view/Integration/Test/Test_Code_Guidelines#User_configurable_variables
Change-Id: I08982748380c54b483383fbcd07a1b6ce617d1fb
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Thu, 3 Mar 2016 13:10:06 +0000 (14:10 +0100)]
Tidied updated test suite
Change-Id: I9ca8ba1794dc4f6117d4f46ff1553ed7a9b1f45a
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Vratko Polak [Tue, 1 Mar 2016 12:21:19 +0000 (13:21 +0100)]
Add odl-cluster-rpc back to akka conf template
This section does not affect Beryllium+,
but fixes Lithium clustering.
As repeatedly stated, this is just a temporary solution,
we need better way of choosing proper template.
Change-Id: Idd1d92be27e500aa85e55db36576efc1d7971ba7
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Jozef Behran [Fri, 26 Feb 2016 10:38:26 +0000 (11:38 +0100)]
Fix variable naming for some immutable variables
A bunch of immutable but not user configurable variables now
need to use snake_case instead of ALL_CAPS due to new
variable naming guidelines [1].
[1] https://wiki.opendaylight.org/view/Integration/Test/Test_Code_Guidelines#Variable_naming
Change-Id: I45f4e3c684c2b2f1905a9ad91afe2cf2fbbb6569
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jozef Behran [Thu, 25 Feb 2016 14:51:31 +0000 (15:51 +0100)]
Fix timeouts for Install/Uninstall Karaf Feature
15 seconds is not enough for some of the features. Even one
minute may be insufficient. Increase thus the default
timeout for these keywords to 3 minutes.
Change-Id: If79ece224becea091d3c42c2ecdac84be2b95861
Signed-off-by: Jozef Behran <jbehran@cisco.com>
Jamo Luhrsen [Thu, 25 Feb 2016 05:43:48 +0000 (21:43 -0800)]
fixing syntax for chaning odl_version in devstack setup
Change-Id: I46db6702957d4b3404cee556255909a2c88a7ff6
Signed-off-by: Jamo Luhrsen <jluhrsen@redhat.com>
Luis Gomez [Wed, 24 Feb 2016 05:34:12 +0000 (21:34 -0800)]
Add topology check in OpenFlow cluster test
Add 2 more switches connecting to first switch candidate instance.
Verify new inventory and topology during owner instance failure and recover.
verify topology when link between switches goes down and up.
Change-Id: I153a24599ce9478bc8e57cd161d46c606aa26bad
Signed-off-by: Luis Gomez <ecelgp@gmail.com>
Vratko Polak [Thu, 25 Feb 2016 17:29:42 +0000 (18:29 +0100)]
Comment out Netconf (netopeer) suite from userfeatures testplan
Change-Id: I511606ca3d4f17d11518ccefa727af220fbf9c75
Signed-off-by: Vratko Polak <vrpolak@cisco.com>