Add TypedReflections

We have multiple places which are accessing annotation-based type
information. This provides a single place to obtain that information.

As we have proper encapsulation, add caches for table/column
versions. Since we are dealing with Version, this class is updated
to be immutable and provide a Range<Version> factory.

Change-Id: I5c2f1d3553392cbd222a1927fdd36564f895475b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Add @NonNull annotation to OvsdbConnectionListener.connected()

Eclipse warns about users overriding unconstrained parameter here,
and we know the instance will not be null. Propagate @NonNull to
the API specification, fixing warnings.

This also fixes similar warnings in other places.

Change-Id: I6cf71c4c0495ba0358157ee12e92b9a7a665b92c
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Add TypedDatabaseSchema

This adds a new entrypoint for performing typed operations on top
of DatabaseSchema. The implementation fowards to TyperUtils, allowing
transition away from those utilities.

OvsdbClientImpl is update to always return TypedDatabaseSchema and
use it to implement its interface.

Change-Id: I6e31f418b2acdebdca01181522b6bb7966061617
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Turn DatabaseSchema into an interface

This moves all the implementation details out into a separate class,
allowing DatabaseSchema interface to be extended with additional
functionality.

Change-Id: Ib7a0124993cb89bb448b3c08bda6c6de0fea1dde
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Make TableSchema/DatabaseSchema immutable

Having schemas completely immutable is beneficial, as we can
propagate invariants properly. This patch makes TableSchema
and DatabaseSchema use immutable maps internally, making sure
the final bit is achieved.

Change-Id: Ie7c9fd4aa9cba7f0af88c8886400e8e5c804d6af
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Split out BaseTypeFactories

Instantiating BaseTypes with complex definition is a prime example
of where either a builder or a factory pattern is appropriate. Given
we are dealing with a few types with very few customizations, this
patch applies the factory pattern to the problem.

This makes all BaseType subclasses properly immutable and has the
benefit of squashing instances to singletons when they are not
otherwise customized.

Change-Id: I86d09324624c1e5b1a73a0be31269f2233fc15b3
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Use singleton BaseType instances for simple definitions

We have two distinct cases for a type definition. It is either
a simple reference without any further constraints or it is
customized in some way.

This patch separates the two codepaths, making the former use
singleton instances.

Change-Id: Ic9e63f2320a429bda076e4ae25eea97e5242a52c
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Split out BaseTypes

This splits out and hides concrete BaseType classes, hiding them
and untangling BaseType.fromJson() instantiation. This makes object
lifecycle more clear, allowing for future optimization.

Change-Id: I60630b3cd7b02d6e9c241d6e666550ae041c96f9
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Cleanup HwvtepConnectionManager.getHwvtepGlobalTableEntry()

Rework the flow of this method to get rid of unneeded null
initializer, making the results in error paths clearer.

Change-Id: If7361c71aefd78ae08a385dcfa7e312ad22f2d5d
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Do not allow DatabaseSchema name/version to be mutated

The setters to do so are not used anywhere, eliminate them, making
it clear these two fields are final.

Change-Id: I817dfb0cc0791c2ad91671b6d5e637c8ae8a0d92
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Do not allow TableSchema columns to be directly set

This makes it obvious that the collection is largely controlled
by TableSchema, bringing it another step closer to being immutable.

Change-Id: Iacb375c3cafbe687966729626fe30a3d08e8f31e
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Refactor ColumnType

This factors out the two known implementations, hiding them from
public view in process. Parsing from JSON is factored out so that
we do not have blank singletons just to hook parsing.

Change-Id: I3dd7b110fb0ca3fe18ef879a2c3d24523fc43419
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Cleanup ColumnSchema

Fields should be private (making it obvious they are final) and
the fromJson() method should be properly generic.

Change-Id: Ic38e4179f1108f41882bd336b3a03b339d435f9c
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Add generated serialVersionUUID to exceptions

Exceptions are serializable and therefore should have a serial
version. Fixes Eclipse warnings pointing this out.

Change-Id: I2cf04412709cf3214b53aa3ac5dc40f332f67295
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Make GenericTableSchema.fromJson() a factory method

This method is called from precisely one place, which is creating
a new object for the call. Turn allocation around, taking a step
towards immutable schemas.

Change-Id: I97c44da72bc3b63260941bcf7d480a9c0c9e3a4a
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Move ObjectMapper to JsonRpcEndpoint

JSON mapping is an implementation detail of JsonRpcEndpoint, hence
move it to that class, allowing us to shed a field and statically
bind to it.

Change-Id: Ie0a314b3aa6c830d9cc5936eb4ec948823ba351c
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Improve schemas population

Fetching the schema is an asynchronous process, which means we could
end up with competing requests overwriting the schema. Make sure to
re-check presence when the resolution process completes.

Change-Id: I0a7076b7a476394d4815f3c8f544be7a9ffaddd8
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Turn JsonRpcEndpoint into a proper OvsdbRPC implementation

The Netty wiring and invocation handling is quite arcane here, relying
mostly on reflection to perform dispatch.

There is only a single direct user JsonRpcEndpoint, which expects
a OvsdbRPC to be exposed from it. Furthermore it is obvious that
JsonRpcEndpoint has a 1:1 correspondence with JsonRpcServiceBinderHandler
and that the invocation context used to pass request processing relies
on Netty Channel instance -- to which the JsonRpcEndpoint is already
bound.

This integrates JsonRpcServiceBinderHandler into JsonRpcEndpoint,
thus making the fusion obvious, eliminating the need to pass Object
context between the two.

Furthermore we eliminate JsonRpcEndpoint.getClient() and opt to make
JsonRpcEndpoint implement OvsdbRPC, providing a very clear connection
between the two and completely eliminating use of reflection in the
process.

Change-Id: I1cfcf784330fdf6e30a55fcaff6f5bd716d987e3
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Reuse ObjectMapper across all connections

ObjectMapper is documented to be safe for sharing, make sure we
reuse a single instance across all connections.

Change-Id: Ia3ebbe1eb4772efaa1f68c0b0d4ee721431325b0
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Use a constant ObjectMapper in UpdateNotificationDeser

Rather than instantiating a fresh mapper for each request,
share a single pre-configured instance.

Change-Id: I231165f82dc950ae9e3adbfb816da123139cc878
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Use proper constant in JsonUtils

We are not using the Mapper here at all, just the writer. Make sure
we retain only the writer and make it a proper constant (to allow
JIT to do constant propagation).

Change-Id: I7df2a6234511c9dbf51509c495a59e642173fffe
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Do not reconfigure ObjectMapper in FutureTransformUtils

This is a shared ObjectMapper, which will be reconfigured on first
access -- we can do that upfront and save some cycles from the fast
path.

Change-Id: Ibdf27ff194350b96ee6aadeaa9725d4cfdc6e9db
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

delete reconciliation support on termination points

Descripion:
There exists chances of stale tunnel-ports left over under following
scenario.

1) DPN connected and Tunnel configuration pushed in by ODL.
2) DPN disconnected and Tunnel Configuation delete by ODL from
network-topology config DS.
3) On DPN connecting back, there exists stale tunnelport port on this
DPN.

Following Changes are done to handle delete reconciliation support.

1) Configure "created_by:odl" on tunnel port in OVS(Port, Interface)
which will be pushed by ovsdb plugin.
2) During reconcilation, compare the termination points available in
config and oper DS and trigger delete of those stale termination points
of Oper DS which are set with "created_by:odl"

JIRA: OVSDB-459

Change-Id: I9356d5ef0af44334c97b6cf3632d4ce9fd2070aa
Signed-off-by: Chetan Arakere Gowdru <chetan.arakere@altencalsoftlabs.com>
Signed-off-by: xcheara <chetan.arakere@altencalsoftlabs.com>

Bump to odlparent-6.0.1/yangtools-4.0.2/mdsal-5.0.4

This picks up latest upgrades and fixes.

Change-Id: I8db358b0af81394a561071d98bffea168dda47cc
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Migrate MockitoJUnitRunner

This migrates from org.mockito.runners.MockitoJUnitRunner to
org.mockito.junit.MockitoJUnitRunner.

Change-Id: I7311c18d01dd51f3ddaa5397c7f61ee3c143cb7f
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Do not use Foo.toString() when logging

This was spotted in a JFR run, where JsonRpcEndpoint would consume
a lot of CPU formatting a logging object. Eliminate calls to toString()
in logging arguments, as the framework will call these when really
needed.

Change-Id: Id077c285074c1fd8012539ab7c144ddc4b286376
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump mdsal to 5.0.3

This bumps mdsal to 5.0.3 to workaround issues with Blueprint
injection.

Change-Id: I848c9220ce06ec7b4af304ab64948bdc30257dbd
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Add a workaround for ARIES-1923

The services here may fail to resolve, causing problems downsteam.
This fixes up the issue by dropping target Java version to 10.

Change-Id: I226ba17183aa202209bbfbbe9ae2b87af7103383
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump mdsal to 5.0.2

This bumps mdsal to 5.0.2 as part of Magnesium MRI.

Change-Id: I93e10e95b7a241d678829afb732e2adef82c1a24
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump odlparent to 6.0.0

This bumps odlparent to 6.0.0 as part of Magnesium MRI.

Change-Id: I084c6a32c4facafd85e3b87c6d8a9a29a33848a8
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Update docs header to Magnesium in the master

Should be Magnesium on master and Sodium on stable/sodium.

Signed-off-by: Anushka Bhandari <anushka16134@iiitd.ac.in>
Change-Id: Ib32db17f852043beadda5932c300b16dbe431e79

Check TP type rather than doing a blind cast

Here we have a blind cast, which does not go through:

[main] WARN org.opendaylight.ovsdb.hwvtepsouthbound.transact.TransactUtils - Failed to add to Transaction Histroy
java.lang.ClassCastException: org.opendaylight.yang.gen.v1.urn.tbd.params.xml.ns.yang.network.topology.rev131021.network.topology.topology.node.TerminationPoint$$$codecImpl cannot be cast to org.opendaylight.ovsdb.schema.hardwarevtep.PhysicalLocator
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.TransactUtils.addLocatorToTransactionHistory(TransactUtils.java:270)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.TransactUtils.createPhysicalLocatorSet(TransactUtils.java:139)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.McastMacsRemoteUpdateCommand.setLocatorSet(McastMacsRemoteUpdateCommand.java:135)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.McastMacsRemoteUpdateCommand.doDeviceTransaction(McastMacsRemoteUpdateCommand.java:93)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.McastMacsRemoteUpdateCommand.doDeviceTransaction(McastMacsRemoteUpdateCommand.java:1)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.AbstractTransactCommand.processDependencies(AbstractTransactCommand.java:125)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.McastMacsRemoteUpdateCommand.onConfigUpdate(McastMacsRemoteUpdateCommand.java:78)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.McastMacsRemoteUpdateCommand.updateMcastMacRemote(McastMacsRemoteUpdateCommand.java:65)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.McastMacsRemoteUpdateCommand.execute(McastMacsRemoteUpdateCommand.java:55)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.TransactCommandAggregator.execute(TransactCommandAggregator.java:73)
at org.opendaylight.ovsdb.hwvtepsouthbound.transact.TransactInvokerImpl.invoke(TransactInvokerImpl.java:41)
at org.opendaylight.ovsdb.hwvtepsouthbound.HwvtepConnectionInstance.transact(HwvtepConnectionInstance.java:132)
at org.opendaylight.ovsdb.hwvtepsouthbound.HwvtepConnectionInstance.transact(HwvtepConnectionInstance.java:104)
at org.opendaylight.ovsdb.hwvtepsouthbound.HwvtepDataChangeListener.updateData(HwvtepDataChangeListener.java:178)
at org.opendaylight.ovsdb.hwvtepsouthbound.HwvtepDataChangeListener.onDataTreeChanged(HwvtepDataChangeListener.java:81)
at org.opendaylight.controller.md.sal.binding.impl.BindingDOMDataTreeChangeListenerAdapter.onDataTreeChanged(BindingDOMDataTreeChangeListenerAdapter.java:43)

Rather than incurring this, perform an instanceof check and debug
ignored data.

Change-Id: Icc3d1abe5afd2b9b60c7632606551904b7d27af7
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Migrate to use ArgumentMatchers

org.mockito.Matchers.any() has been deprecated, use its newer
equivalent.

Change-Id: I46e4dd8000ab5a047da8414b5e9707b5f5e3b056
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump mdsal to 4.0.6

This picks up latest fixes.

Change-Id: I64e1f277bd00e132f40647eb0a2e7fc00eb56705
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump odlparent to 5.0.2

This bumps odlparent to pick up latest references.

Change-Id: I73af7ad69f686cb5bacc6ccfe0f57d226a43994b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Fix checkstyle

This fixes up checkstyle violations.

Change-Id: I7ee4391b58eef76954157deb11b611a3837d0960
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump mdsal to 4.0.4

This brings in latest fixes.

Change-Id: I59f89ef4fe76ef16fcdc9b59f8d9b50aa0849847
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump versions by x.y.(z+1)

Signed-off-by: jenkins-releng <jenkins-releng@opendaylight.org>
Change-Id: I876601792e0d95e76362d9fed8c1f2c9b0c4d553

Bump mdsal to 4.0.3

This bumps mdsal to 4.0.3, bringing in latest fixes and improvements.

Change-Id: Ie8e03aab1f31edf80bc9f1d7896c334c9d83be8d
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump odlparent to 5.0.1

This picks up the latest upgrades.

Change-Id: If6006a94197839a9718cd5eae16b0a025913af93
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Optimization to read the HwvtepNode from the cache.

JIRA: OVSDB-485

Currently in the HwvtepSouthbound plugin reads the HwvtepNode from the datastore.
Reading the HwvtepNode from the datastore ineffecient, even the same info
is present in HwvtepSouthbound plugin cache.

This review is avoid the HwvtepNode reads from the datastore and get it from the cache.
If the node is not present in the cache, then read from datastore.

Signed-off-by: Chandra Shekar S <chandra.shekar.s@ericsson.com>
Change-Id: I1f15ff31b195a79a1725e63882512b87d40c5838

OVSDB-486 Fix Null Pointer Exception in OVSDB

Fix for Null Pointer Exception

Change-Id: I93227ab66df4529a9b8c127c4b41096335424cc3
Signed-off-by: Nobin Mathew <nobin.mathew@ericsson.com>

Avoid Bulk Topology Read During Reconciliation.

Description:
During Reconciliation, The full config network-topology DS is queried
which is not efficient. Currently, only br-int is managed in
network-toplogy config DS(by genius) which will only be resynced.

Changes are done to query only the list of bridges for a given connected
switch (configurable vi param - bridges-reconciliation-inclusion-list) from
topology config instead of querying the full topology from config DS and
filtering the bridges specific to connected Switch.Following
implementation been provided(to keep the existing behaviour)

(1) Both "bridge-reconciliation-inclusion-list" and "bridge-reconciliation-exclusion-list" are empty.
    it means it will keep the default behavior of reconciling on all bridges.
(2) Only "bridge-reconciliation-inclusion-list" has list of bridge.
    than plugin will only reconcile specified bridges.
(3) Only "bridge-reconciliation-exclusion-list" has list of bridge.
    than plugin will reconcile all the bridge, except excluding the specified bridges.
(4) Both bridge-reconciliation-inclusion-list and bridge-reconciliation-exclusion-list has bridges specified.
    this is invalid scenario, so it should log the warning saying this is not valid configuration,
    but plugin will give priority to "bridge-reconciliation-exclusion-list" and reconcile all the bridges
    except the one specified in the exclusion-list.

JIRA: OVSDB-459

Change-Id: Ife8055bcc10bcb6deb2fd8e0b5444bdf194f1460
Signed-off-by: Chetan Arakere Gowdru <chetan.arakere@altencalsoftlabs.com>

Karaf cli command to display the hwvtep cache data

JIRA: OVSDB-482

The hwvtep configuration data is pushed to the tor by the hwvtep plugin.
The hwvtep plugin has the caches (config and operational caches) which holds
the hwvtep configuration. The hwvtep configuration push to the tor is also
dependent on the data that is present in caches.
For ex : To push the vlanbinding data to the PhysicalPort , the operational
cache is checked first, if present in the operational cache get the uuid of
that PhysicalPort and the send/push the vlanbinding to the tor.

Today it is hard to know/check what is present in the cache. In case if there
any issue in updating the cache, it is hard to debug/analyse the issue as no way
to know what is present in the cache.

If we have mechanism/cli command which gives the information of the hwvtep
caches, it will be very helpful to debug or analyze such cache missing issues.

This is for adding the karaf cli command which displays the hwvtep caches contents.
This command has two options :
1. Hwvtep nodeid as the input. Displays cache information of input hwvtep node.
2. Without any input. Displays the cache information of all hwvtep nodes.

Change-Id: I4301a784d83aa28f5ef4737fc44d052f323e11df
Signed-off-by: Chandra Shekar S <chandra.shekar.s@ericsson.com>

Optimizatoin to get the nodeIid from the DeviceOperCache in HwvtepPhysicalPortUpdateCommand

JIRA: OVSDB-482

Description :
Currently in HwvtepPhysicalPortUpdateCommand while updating the terminationpoint the
nodeIid/switchIid is retrieved from the incoming event data, if not present in the
incoming event data, it is retrieved from the DataStore by looking/looping at all
switches for that node which is ineffecient. The logic has chances of throwing NPE when
there are no PhisicalSwitches present for that node in the DataStore.

The change is to retrieve the nodeIid from the deviceOperCache , if not present get it from the
DataStore. The change also contains the handling of the NPE while retrieving the data from DataStore

Change-Id: Ifbfad7eb9cdac1c2eedb29ea9eb2a303ac8a9a3c
Signed-off-by: Chandra Shekar S <chandra.shekar.s@ericsson.com>

Hwvtep TransactionHistory improvements.

JIRA: OVSDB-481

The change in this review is to improve the format of the Hwvtep TransctionHistory hwvtep:txlog command
1. The hwvtep:txlog output is not logging/printing the date correctly. The date that is getting
printed is the date at which the command is executed , not the date at the transaction is looged.
Change has the fix for date issue.

2. The hwvtep:txlog output format is printing the complete instance identifier information of the nodes.
  making the command output clumsy and hard to read.
  ex :
    Printing for iid KeyedInstanceIdentifier{targetType=interface org.opendaylight.yang.gen.v1.urn.tbd.
    params.xml.ns.yang.network.topology.rev131021.network.topology.topology.Node, path=
    org.opendaylight.yang.gen.v1.urn.tbd.params.xml.ns.yang.network.topology.rev131021.NetworkTopology,
    org.opendaylight.yang.gen.v1.urn.tbd.params.xml.ns.yang.network.topology.rev131021.network.topology.
    Topology[key=TopologyKey{_topologyId=Uri{_value=hwvtep:1}}],org.opendaylight.yang.gen.v1.urn.tbd.
    params.xml.ns.yang.network.topology.rev131021.network.topology.topology.Node[key=NodeKey{_nodeId=
    Uri{_value=hwvtep://uuid/3b0fb1b3-6449-4fa7-b6ac-0db437fe144e}}]]}
Update the output format only to print the node id : "hwvtep://uuid/3b0fb1b3-6449-4fa7-b6ac-0db437fe144e"

3. Update the TransctionHistory for the missing tables/objects like LogicalSwitch and Physical Port etc.
4. Added more logging of Transctions.

Change-Id: Ia5e8984fb414944820140acae69dd64104ee9619
Signed-off-by: Chandra Shekar S <chandra.shekar.s@ericsson.com>

Spec for OVSDB reconciliation

JIRA: OVSDB-459
Change-Id: I913814e61a2bfc1a10bdbc4c65b3108208108cd1
Signed-off-by: Nobin Mathew <nobin.mathew@ericsson.com>
Signed-off-by: Chetan Arakere Gowdru <chetan.arakere@altencalsoftlabs.com>

Remove use of StringBuffer

StringBuilder is a faster replacement, use that instead of
StringBuffer.

Change-Id: I5baafddeefff469796cd33dea7056cc0d6a5f6c5
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Update docs/ovsdb-user-guide.rst and README

odl-ovsdb-openstack has been moved to netvirt and changed as
odl-netvirt-openstack, so update it to reflect current status.

JIRA: OVSDB-480

Change-Id: Ied1268898d0adc63894250b9733a974140bc31c4
Signed-off-by: Yi Yang <yangyi01@inspur.com>

Bump mdsal to 4.0.2

This picks up the latest fixes.

Change-Id: I60557d40e4ae849ea73f3232f20f7032205dbe65
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump mdsal to 4.0.1

This bumps mdsal to 4.0.1.

Change-Id: I23245cdccf757cf2c1c19a6774cc5fbb71382343
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Allow SouthboundUtils to work with MdsalUtils

SouthboundUtils are being used by downstreams, who want to migrate
to MD-SAL APIs. Allow SouthboundUtils to work with both
ControllerMdsalUtils and MdsalUtils.

Change-Id: Iab6e48c060942e56db111755c08cd15206d38695
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Cleanup OvsdbDataTreeChangeListener

This patch restructures the logic a bit, so that it does not invoke
getData{After,Before}() multiple times, fixing Eclipse warnings and
improving performance.

Furthermore augmentation handling is refactored so that the two
cases (node/bridge) are handled in an if cascade, preserving type
safety.

Finally updateConnections() is refactored to use a switch statement
instead of multiple if conditions, which shows that delete of the
entire node are not handled -- a FIXME is placed there.

Change-Id: If5663ebd2944d7fe83c4928b96cedc0db0f90e19
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump mdsal to 4.0.0

This bumps mdsal to 4.0.0 as part of Sodium MRI.

Change-Id: I7c4c8956623072e402371147e58e9cc2a2e4bf78
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump odlparent to 5.0.0

This bumps odlparent to 5.0.0 as part of Sodium MRI, also migrates
simple use of JSR305 to JDT.

Change-Id: Ie115a811e9f1a13e0ed4d3e1008270b589b4bdcc
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Fix logging arguments

These are issues pointed out by updated SpotBugs, fix them up.

Change-Id: I2e740a1e23b96cbe12e4a49b0ca306876c5a713b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Remove use of deprecated Guava methods

Pass down MoreExecutors.directExecutor() to future-proof the code.

Change-Id: Iaf847f2dbcce63bac685d5a22a35fc9b8085d06d
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Upgrade powermock to 2.0.0

This upgrades powermock to version 2.0.0, also updating tests to
match nulls being returned through mocks.

JIRA: OVSDB-476
Change-Id: I54beb2db411cff0e126c8e8208ff9ee1530437ab
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Fix static mocking in TransactUtilsTest

Change static mocks to doReturn().when(...) mocking style, which
allows them to properly work with Mockito 2.

Change-Id: Ie5b2f0223fd67b942a5c77cd56f5ccaa4fa870e1
JIRA: OVSDB-476
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Fix SouthboundUtilTest

Use PowerMockito.doReturn() to correctly mock static method invocation,
enabling the test to pass.

Change-Id: I7d999ed1ca3720877ea8ee9e6d59a267604f3409
JIRA: OVSDB-476
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Remove unnecessary stubbing

This fixes interop with Mockito 2, allowing the suite to pass
further.

Change-Id: I22c65f2b3a0f732c2b6bea6554ad4d75aa520f82
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Fix OvsdbConnectionManagerTest

Fix mock returns to use proper types, so this works with Mockito2,
too.

Change-Id: I277ade6a61978bd829c437fb9a90fc91d58045f2
JIRA: OVSDB-476
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Fix OvsdbManagersRemovedCommandTest

Use proper mocking to return non-null values when intercepting
methods.

Change-Id: Ic70d6696ce869d91d8de72859130f758d68de830
JIRA: OVSDB-476
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Fix OvsdbConnectionInstanceTest

Remove unnecessary stubbing and fix any()/mock() confusion.

JIRA: OVSDB-476
Change-Id: I7621f67e3af5d1191c0a51e3356d257263991202
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Fix OpenVSwitchUpdateCommandTest

There is some unnecessary stubbing, remove it to ease migration.

JIRA: OVSDB-476
Change-Id: Ie8222eaa524a5fabb9b054e6076a3377708c7248
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Remove ProtocolUpdateCommandTest

This test is not verifying anything, remove it.

Change-Id: I81ebccd450d888adfd66420f90cfbcd907996c4d
JIRA: OVSDB-476
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Remove OvsdbControllerRemovedCommandTest

This test is using partial mocking to the point of not checking
anything real and actually verifies incorrect internal method
signatures. Remove it.

JIRA: OVSDB-476
Change-Id: I04e67ff6937e84cdc55e1347e407596df8605c99
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Use JDT annotations instead of JSR-305

This reduces our dependencies on javax.annotation, which end up
causing problems with Java 11.

Change-Id: Ied3be5b9677bef722c008b2828950bb12ae84678
Signed-off-by: Stephen Kitt <skitt@redhat.com>

Do not workaround MDSAL errors

This code masked an MD-SAL failure to account for UNMODIFIED nodes,
which luckily had no effect. Upstream has this fix in the pipeline,
this patch removes the workaround.

Change-Id: I150a50c46e3272a0b099030f4424628ed983edff
JIRA: MDSAL-422
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Reduce use of powermockito

Remove unneeded supressions and mocking, with judicious reuse
of existing methods, @VisibleForTesting and similar.

Change-Id: If6aff32331d2a50752f5cd2ef981abe4d49f0af5
JIRA: OVSDB-476
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Remove useless tests

These tests are not asserting anything at all, remove them to reduce
static mocking proliferation for no value whatsoever.

Change-Id: I53b21da97d7b9c7cf6e64543e6fe73f2c08267e9
JIRA: OVSDB-476
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Reduce use of powermockito

Rather than mucking with powermock, use Mockito's partial mocking
on a real instance of HwvtepConnectionManager.

Also proper application of @VisibleForTesting and ovsdbClient mocking
allows us to not mock HwvtepConnectionInstance at all.

We also have a simple utility for setting final fields, use that
to reset Operations static field -- allowing us to not use powermock
in DataChangeListenerTestBase at all.

JIRA: OVSDB-476
Change-Id: I6c321c0455fb886dc0029b2a321f73eb3ba150d1
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Remove odlparent overrides

This removes unneeded overrides of odlparent-exported artifacts.

Change-Id: I7ee16354e180609babc65b1fb0a80d005489355b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Switch to spotbugs plugin

findbugs does not work JDK9+, switch to spotbugs.

Change-Id: I2672369099c7298c67f9d50acec2bbdb882fb1d3
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Upgrade powermock utils/

This upgrades powermock and fixes up warnings around mockito
being deprecated.

Change-Id: I74fe7b16b223ee9bab966478f6a653281634b8d7
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump versions by x.y.(z+1)

Change-Id: I21c38dd91233b8652a2a15d0cc9badb08c9e1931
Signed-off-by: jenkins-releng <jenkins-releng@opendaylight.org>

Remove duplicate artifact declarations

This fixes a few maven warnings about artifacts being declared
multiple times.

Change-Id: Ib6a7a31931883009868f6e4150a8cb0ed5718fd6
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

utils: migrate to the mdsal DataBroker

This duplicates controller-based classes temporarily, so that we can
migrate hwvtepsouthbound and southbound separately.

Change-Id: I1ed7c137423deb149320df86f0abc090e1182df5
Signed-off-by: Stephen Kitt <skitt@redhat.com>

library: migrate to the mdsal DataBroker

Change-Id: I0540965fe40521dd2d18e2206fb2bb874e2f1e72
Signed-off-by: Stephen Kitt <skitt@redhat.com>

Add missing annotation dependencies for Java 11

Change-Id: I552dc74b2bb6832cc5419f309470a43430294443
Signed-off-by: Stephen Kitt <skitt@redhat.com>

use annotations instead of XML for Blueprint

JIRA: OVSDB-474
Change-Id: Ie40dacf4b4a745d77d83bdc7e1b95fecad4fd276
Signed-off-by: Harshini <hm@luminanetworks.com>
Signed-off-by: Michael Vorburger <vorburger@redhat.com>

Bump to mdsal 3.0.6

Change-Id: Id4a938a3e8d8669b972df17ddf1d76f91ca686d0
Signed-off-by: Stephen Kitt <skitt@redhat.com>

Bump mdsal to 3.0.5

This aligns with odlparent-4.0.9 and yangtools-2.1.8.

Change-Id: I1800398e0f1cfbfc263164ea6eff878423d31b98
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump to odlparent 4.0.9

Change-Id: Iabeeaae5001bb0fba3c84d22cedd82444b60f7b8
Signed-off-by: Stephen Kitt <skitt@redhat.com>

Bump mdsal to 3.0.4

This realigns odlparent/yangtools/mdsal and brings in the latest
improvements.

Change-Id: I6db0663fcf132b9e01839c2cc14dc611ea75aa60
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump odlparent to 4.0.8

Change-Id: Ieee47dc9a810d99311bb1bfdeedc71655a612cab
Signed-off-by: Stephen Kitt <skitt@redhat.com>

Fix controller for ptcp manager

If a ptcp manager is set on OVS, we configure
controller through conneciton info. There are two
issues with this code today:

1. Doesn't work well for IPv6.
2. Doesn't check if connection is to same port
   as configured in ptcp manager.

So we end up with an improper controller configured
basde on ptcp even if we already configured correct
ones from manager entries.

JIRA: OVSDB-475

Change-Id: Ic69be0a9a6f2fd96085b60cdc3ccb9f1022b54a8
Signed-off-by: Vishal Thapar <vthapar@redhat.com>

Migrate NumberUtils.isNumber()

This method is deprecated, use its replacement, isCreatable().

Change-Id: I47a6dc48e5da306f8a1c2fa2e0a19ca034756c28
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Remove use StringUtils

Guava provides equivalent functionality, use that instead of
commons-lang3.

Change-Id: Ide5fb6606ea51d6a65518e5e57422b11c259d86a
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump mdsal to 3.0.3

This brings in latest fixes and alignes with odlparent-4.0.5.

Change-Id: I4626267cb5d322378e87b82045dd4f080a747e35
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Bump to odlparent 4.0.5

Change-Id: Ib2f5fd0e462e8a1efdefc2e4e94acc109c688c19
Signed-off-by: Stephen Kitt <skitt@redhat.com>

Use targeted node type for bridge name search

The IdentifiableItem we're looking for when checking for a bridge name
is Node; this changes isBridgeOnOvsdbNode() to look for that
directly. If also changes toString() to a match-specific method to
avoid the temptation to use it for other purposes.

Change-Id: Ib20e2c909098af9e0643f12805d98c5f2a66b651
JIRA: OVSDB-470
Signed-off-by: Stephen Kitt <skitt@redhat.com>

reduce object allocation in SouthboundUtils.isBridgeOnOvsdbNode()

By avoiding any unnecessary toString() usage, but unfortunately not
being able to complete avoid it, without further/bigger/riskier changes.

This should help to reduce GC churn.

JIRA: OVSDB-470
Change-Id: I86cdba3c138a79b428acf960a2ddb035f2878ca9
Signed-off-by: Michael Vorburger <vorburger@redhat.com>

Bump to mdsal-3.0.2

This bumps the dependency to 3.0.2.

Change-Id: I4ab38f4dce8983e310430328c4367c732a5af3ae
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Fixes overwrite when adding termination point

A race condition was observed where OVSDB had already added the VTEP TP
to br-int, and following this Netvirt ELAN Manager added the patch port
(br-ex-patch) to br-int. Upon ELAN manager adding this using
SouthboundUtils addTerminationPoint, the previous VTEP was deleted. This
change modifies the addTerminationPoint to use a merge in MDSAL rather
than a put.

JIRA: OVSDB-472

Change-Id: I2afa7dc33ef915856774f431dd4c3d9bcd029155
Signed-off-by: Tim Rozet <trozet@redhat.com>

Log the address and port on bind errors

This ensures the address and port are available in the logs when we
fail to bind.

Change-Id: Iae6a6fd4d6f516d5dbd562a335d081181cff3b17
JIRA: OVSDB-473
Signed-off-by: Stephen Kitt <skitt@redhat.com>

Fix ovsdb compilation

southbound-impl relied on getting its dependencies transitively:
- sal-binding-api via aaa
- mdsal-dom-api via sal-core-api

Fix this up by specifying correct dependencies.

Change-Id: Iba60e8978c29a82f77c42d7ce1ee6052613fdcea
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Inject the certificate manager in the constructor

This simplifies wiring in other contexts (e.g. with Blueprint
annotations or Guice).

Change-Id: If76dae1ec26d3ef3e3488788b6a5b77d99bb70a1
Signed-off-by: Stephen Kitt <skitt@redhat.com>

Upgrade to Neon base platform

This reverts the PowerMock and Mockito upgrades for the time being.

Change-Id: I46d935fed3ef44dede188ffd25295586010cc8a2
Signed-off-by: Vishal Thapar <vthapar@redhat.com>
Signed-off-by: Stephen Kitt <skitt@redhat.com>
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>

Update docs header to Neon on master

This version name gets rendered into the main header of OVSDB's docs.

Now that these docs have moved from the docs repo to project-owned
repos, projects need to update this version themselves.

Should be Neon on master and Fluorine on stable/fluorine.

Change-Id: I82a4bffc06a894bbc643c496c0471f3fe413a280
Jira: DOCS-112
Signed-off-by: Daniel Farrell <dfarrell@redhat.com>