genius.git
7 years agoMerge "Handle duplicate tunnel on multiple dpns"
Vishal Thapar [Thu, 22 Dec 2016 01:58:18 +0000 (01:58 +0000)]
Merge "Handle duplicate tunnel on multiple dpns"

7 years agoMerge "Update spec template"
Vishal Thapar [Wed, 21 Dec 2016 07:58:06 +0000 (07:58 +0000)]
Merge "Update spec template"

7 years agoHandle duplicate tunnel on multiple dpns 38/49638/1
Vishal Thapar [Tue, 20 Dec 2016 11:10:12 +0000 (16:40 +0530)]
Handle duplicate tunnel on multiple dpns

If same tunnel port name shows up on multiple dpns, it is wrongly assumed
to be a case of VM Migration, which only applies to VM ports. In case we
get any such ports, first check on which DPN user configured it. Retain
interface-state entry for the port that is create on the DPN as configured
by user.

Change-Id: I95dab0a36fd1d033747fd887a4fea4a648da04aa
Signed-off-by: Vishal Thapar <vishal.thapar@ericsson.com>
7 years agoDataStoreJobCoordinator super minor LOG related performance fix 90/49490/4
Michael Vorburger [Fri, 16 Dec 2016 17:01:18 +0000 (18:01 +0100)]
DataStoreJobCoordinator super minor LOG related performance fix

LOG statements, especially trace(), should not do object creation, and
in this particular case, it seems not necessary to transform the Set to
an Array to a String... Set has a suitable toString, so not sure why
this was thought to be necessary; it should be just fine like this.

PS: Found while staring at DataStoreJobCoordinator for async test
infra.. I still can't complete make sense of what this class really
does! (Thus nor how to sanely make it testable.)

Change-Id: I27593b5ca0a5afe1784be23bc49b3fc231aaee51
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoUpdate spec template 58/49558/1
Vishal Thapar [Mon, 19 Dec 2016 15:06:06 +0000 (20:36 +0530)]
Update spec template

Refer: https://git.opendaylight.org/gerrit/#/c/49459/

This also adds ToC to existing specs.

Change-Id: I26322d3b288fd8e422b9ee504d72ccdc6ced24c4
Signed-off-by: Vishal Thapar <vishal.thapar@ericsson.com>
7 years agoInterfaceManagerConfigurationTest reformatting etc. so no warnings 03/49303/3
Michael Vorburger [Tue, 13 Dec 2016 14:32:04 +0000 (15:32 +0100)]
InterfaceManagerConfigurationTest reformatting etc. so no warnings

Change-Id: Ifa17867b095f600f0e14b354d17d5f3d3f5ac36f
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoMerge "ResourceBatchingManager code clean up and fixed bad LOG"
David Suarez [Sun, 18 Dec 2016 23:10:48 +0000 (23:10 +0000)]
Merge "ResourceBatchingManager code clean up and fixed bad LOG"

7 years agointerfacemanager POM fixed - no *-impl bundles 95/49495/1
Michael Vorburger [Fri, 16 Dec 2016 17:48:14 +0000 (18:48 +0100)]
interfacemanager POM fixed - no *-impl bundles

Not sure what historically happened here, but obviously the
implementation bundles of other modules should never be direct runtime
dependencies.. only their *-api.  And for component tests, some impl may
be <scope>test dependencies.

Change-Id: I44afb68fa4db3d795cdb045b05f8c16d4a369d0d
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoResourceBatchingManager code clean up and fixed bad LOG 91/49491/2
Michael Vorburger [Fri, 16 Dec 2016 17:11:29 +0000 (18:11 +0100)]
ResourceBatchingManager code clean up and fixed bad LOG

PS: Found while staring at ResourceBatchingManager for async test
infra.. I still can't complete make sense of what this class really is
supposed to be good for, or what problem it really solves! (Thus nor how
to sanely make it testable.)

Change-Id: Ica12a5695647d7d228cbd4e885ee38943f91e642
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoMerge "Cleanup for constants and other stuff."
Faseela K [Fri, 16 Dec 2016 13:53:33 +0000 (13:53 +0000)]
Merge "Cleanup for constants and other stuff."

7 years agoDataStoreJobCoordinator fixed broken LOG.warn of root cause exception 35/49435/2
Michael Vorburger [Thu, 15 Dec 2016 14:12:28 +0000 (15:12 +0100)]
DataStoreJobCoordinator fixed broken LOG.warn of root cause exception

unless there is a very particular reason why this is done in an unusual
weird style, which does not properly show the Exception root cause in
logs (noticed while investigating a failure of the
InterfaceManagerConfigurationTest), it's wrong.

Change-Id: Ic42b4c79b9e7afe69eb4e4e6c633bb899ed3b74c
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoCleanup for constants and other stuff. 23/48823/6
David Suarez [Wed, 30 Nov 2016 10:28:07 +0000 (11:28 +0100)]
Cleanup for constants and other stuff.

- Declare constants properly.
- Log exceptions.
- Local variables.
- Basic formatting.
- Replace Guava's Preconditions by Objects'
- Remove catching of NPE
- Adding more constants to improve readability

There is still a NPE catching that will be removed in a later patch in
which I'll simplify the way in which we call the RPCs (see line
204-206).

Change-Id: I5a2173a39245e7d65469d84cb2a71ef9e21fe1be
Signed-off-by: David Suarez <david.suarez.fuentes@ericsson.com>
7 years agoMerge "Genius: Neutron Qos DSCP Rule support"
Faseela K [Fri, 16 Dec 2016 08:31:50 +0000 (08:31 +0000)]
Merge "Genius: Neutron Qos DSCP Rule support"

7 years agoMerge "Add CacheBridgeRefEntryListener to provider"
Michael Vorburger [Thu, 15 Dec 2016 16:03:12 +0000 (16:03 +0000)]
Merge "Add CacheBridgeRefEntryListener to provider"

7 years agoAdd CacheBridgeRefEntryListener to provider 31/49331/4
Vishal Thapar [Wed, 14 Dec 2016 07:22:39 +0000 (12:52 +0530)]
Add CacheBridgeRefEntryListener to provider

Change-Id: I9aae46b46ed7da3a820ca97521d95af56bf25385
Signed-off-by: Vishal Thapar <vishal.thapar@ericsson.com>
7 years agoflow entries for multiple subports not getting created 05/49405/2
Faseela K [Thu, 15 Dec 2016 06:08:44 +0000 (11:38 +0530)]
flow entries for multiple subports not getting created

Description :Update of vlanmember interfaces is causing unnecessary churn
of deleting and adding back all the interfaces unnecessarily, even if
there is no attribute which is modified.

Change-Id: Ib2d3be95ce48e2d971b27a8ff6e11d113be532ef
Signed-off-by: Faseela K <faseela.k@ericsson.com>
7 years agoMerge "InterfaceManagerConfigurationTest remove BridgeInterfaceEntryBuilder"
Faseela K [Wed, 14 Dec 2016 12:08:24 +0000 (12:08 +0000)]
Merge "InterfaceManagerConfigurationTest remove BridgeInterfaceEntryBuilder"

7 years agoGenius: Neutron Qos DSCP Rule support 84/49084/7
P.Poovizhi [Wed, 7 Dec 2016 12:09:27 +0000 (17:39 +0530)]
Genius: Neutron Qos DSCP Rule support

Adds QoS as a new service and a new table is created for QOS DSCP marking.
DSCP set action to tos field of the IP header is defined newly.

Testing procedure:
1. When a port is associated with QOS policy having DSCP rule. OF is
polpoluated for table 90.

Spec Review Link:
https://git.opendaylight.org/gerrit/#/c/48949/

Result:
- traffic flows with the DSCP value configured in the port.

Change-Id: I52057973bae295ec2e5ea4aec392b4ddf906f1bb
Signed-off-by: P.Poovizhi <poovizhi.p@ericsson.com>
7 years agoMerge "InterfaceManagerTestModule LOG InterfacemgrProvider close() Exceptions"
Faseela K [Wed, 14 Dec 2016 07:29:04 +0000 (07:29 +0000)]
Merge "InterfaceManagerTestModule LOG InterfacemgrProvider close() Exceptions"

7 years agoInterfaceManagerTestModule LOG InterfacemgrProvider close() Exceptions 24/49324/1
Michael Vorburger [Tue, 13 Dec 2016 22:22:51 +0000 (23:22 +0100)]
InterfaceManagerTestModule LOG InterfacemgrProvider close() Exceptions

Change-Id: I1099c834024d8deb5e230e7a656bce5ed7c5a6e3
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoUse platform-specific line separators 09/49309/1
Stephen Kitt [Tue, 13 Dec 2016 16:26:12 +0000 (17:26 +0100)]
Use platform-specific line separators

This should allow the XtendBeanGenerator-based tests to run on Windows
too.

Change-Id: I5c9e5d79405f8fb9a1f5035ad5a259eb9342614e
Signed-off-by: Stephen Kitt <skitt@redhat.com>
7 years agoInterfaceManagerConfigurationTest remove BridgeInterfaceEntryBuilder 04/49304/1
Michael Vorburger [Tue, 13 Dec 2016 14:33:54 +0000 (15:33 +0100)]
InterfaceManagerConfigurationTest remove BridgeInterfaceEntryBuilder

apparently it's un-used - or is this a mistake, and it should be used?

Change-Id: Id241bec009aadc5f7f20c65396c06b1540ea3054
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoAdd isIpInSubnet utility API to NwUtil 57/49257/2
Miguel Perez [Mon, 12 Dec 2016 17:01:26 +0000 (18:01 +0100)]
Add isIpInSubnet utility API to NwUtil

 + That method is extracted from VpnUtil (netvirt). NwUtil.java is
   a more suitable place

Change-Id: Ie999d09aec188fb861254fc985a53a9bd925ac68
Signed-off-by: Miguel Perez <francisco.miguel.perez@ericsson.com>
7 years agoMerge "Fix the JMXAlarmAgent"
Michael Vorburger [Mon, 12 Dec 2016 15:08:51 +0000 (15:08 +0000)]
Merge "Fix the JMXAlarmAgent"

7 years agoMerge "Bug 7270 Duplicate remote Mcast mac entry in TOR ."
Faseela K [Mon, 12 Dec 2016 08:19:03 +0000 (08:19 +0000)]
Merge "Bug 7270 Duplicate remote Mcast mac entry in TOR ."

7 years agoFix the JMXAlarmAgent 18/49118/2
David Suarez [Wed, 7 Dec 2016 22:17:30 +0000 (23:17 +0100)]
Fix the JMXAlarmAgent

- Declare logger properly.
- More generic interface.
- Fix lifecycle: registration and de-registration.
- Fix logging of exceptions.
- The ItmTunnelEventListener can still work even if the system couldn't
  initialize the JMX agent properly.
- Cleanup and formatting.

Change-Id: Iecea96b426c1e7b30868eebe124c34427bc06f9b
Signed-off-by: David Suarez <david.suarez.fuentes@ericsson.com>
7 years agoEnhancing DataStoreJobCoordinaor logs 16/49216/2
Tomer Pearl [Sun, 11 Dec 2016 11:31:17 +0000 (13:31 +0200)]
Enhancing DataStoreJobCoordinaor logs

In scale testing I had encountered jobs that took a long time to compelete,
between 1 and 10+ seconds.

Adding a warm log line for every job that takes more then 1sec to complete,
in order to help us in the future to detect possible bottlenecks
or just plain bugs.

Change-Id: Id65e00f925267bf1a30112804322d9b359e308eb
Signed-off-by: Tomer Pearl <tomer.pearl@hpe.com>
7 years agoMerge "Bug 7315: switch ActionOutput to Uri"
Sam Hague [Sat, 10 Dec 2016 02:24:16 +0000 (02:24 +0000)]
Merge "Bug 7315: switch ActionOutput to Uri"

7 years agoBlueprint migration for ITM 17/48617/10
Edwin Anthony [Mon, 5 Dec 2016 09:30:05 +0000 (15:00 +0530)]
Blueprint migration for ITM

change-Id: If4a0a852893e2b6d88ec59e0ac88d0af7c2bb3ce
Signed-off-by: Edwin Anthony <n.edwin.anthony@ericsson.com>
7 years agoMerge "Introduce DataStoreJobCoordinator counters"
Sam Hague [Thu, 8 Dec 2016 17:59:45 +0000 (17:59 +0000)]
Merge "Introduce DataStoreJobCoordinator counters"

7 years agoBug 7315: switch ActionOutput to Uri 35/49135/3
Stephen Kitt [Thu, 8 Dec 2016 09:21:16 +0000 (10:21 +0100)]
Bug 7315: switch ActionOutput to Uri

It turns out the output action port really is a string, interpreted by
Uri; so the appropriate type is probably Uri instead of long (or
perhaps String if we want to avoid Uri in the external API).

Change-Id: Ia55734bb4c7656b18c76de123717ea148e8d882f
Signed-off-by: Stephen Kitt <skitt@redhat.com>
7 years agoIntroduce DataStoreJobCoordinator counters 87/48987/3
Olga Schukin [Mon, 5 Dec 2016 14:32:03 +0000 (16:32 +0200)]
Introduce DataStoreJobCoordinator counters

The changes related to Bug 7236: System unable to handle high
rate of src mac learning packet-ins

Change-Id: Ie051e7d3bc5aca15a06e3e2aa42a75dc786f252c
Signed-off-by: Olga Schukin <olga.schukin@hpe.com>
7 years agoCleanup unwanted exceptions in interfacemanager 27/49027/2
Faseela K [Tue, 6 Dec 2016 12:21:57 +0000 (17:51 +0530)]
Cleanup unwanted exceptions in interfacemanager

Description : Fixing NumberFormatExceptions thrown by interfacemanager
when nodeConnectorId comes with reserved port keywords, for eg :
openflow:1:LOCAL instead of Integer port numbers

Change-Id: Ifa5b6d4924fe400345c8ac6ed71c1a7460191e29
Signed-off-by: Faseela K <faseela.k@ericsson.com>
7 years agoMerge "Bug 7220 :port updates are not getting reflected in Table 220"
Sam Hague [Wed, 7 Dec 2016 13:53:50 +0000 (13:53 +0000)]
Merge "Bug 7220 :port updates are not getting reflected in Table 220"

7 years agoBug 7220 :port updates are not getting reflected in Table 220 31/49031/3
Faseela K [Tue, 6 Dec 2016 13:26:31 +0000 (18:56 +0530)]
Bug 7220 :port updates are not getting reflected in Table 220

Description : services-info DS was not getting cleared for interface-state removals
for default egress service, this will have impact when there is a change in the underlying port
number for the interface being bound

Change-Id: I63a105c673dba9f1e1d8e2a6e8b1c6d7e02cf650
Signed-off-by: Faseela K <faseela.k@ericsson.com>
7 years agoMerge "New match Reg4 type and temporary SMAC table definitions"
Vishal Thapar [Wed, 7 Dec 2016 10:41:43 +0000 (10:41 +0000)]
Merge "New match Reg4 type and temporary SMAC table definitions"

7 years agoMerge "Refactor ActionType.nx_conntrack"
Faseela K [Tue, 6 Dec 2016 07:42:23 +0000 (07:42 +0000)]
Merge "Refactor ActionType.nx_conntrack"

7 years agoMerge "Deprecate ActionType.goto_table"
Faseela K [Tue, 6 Dec 2016 07:41:48 +0000 (07:41 +0000)]
Merge "Deprecate ActionType.goto_table"

7 years agoMerge "Refactor push/pop MPLS/PBB/VLAN actions"
Faseela K [Tue, 6 Dec 2016 07:41:41 +0000 (07:41 +0000)]
Merge "Refactor push/pop MPLS/PBB/VLAN actions"

7 years agoMerge "Migrate ActionType.output to ActionOutput"
Faseela K [Tue, 6 Dec 2016 05:49:26 +0000 (05:49 +0000)]
Merge "Migrate ActionType.output to ActionOutput"

7 years agoMerge "Migrate ActionType.group to ActionGroup"
Faseela K [Tue, 6 Dec 2016 05:49:17 +0000 (05:49 +0000)]
Merge "Migrate ActionType.group to ActionGroup"

7 years agoNew match Reg4 type and temporary SMAC table definitions 34/48834/5
Olga Schukin [Wed, 30 Nov 2016 14:29:43 +0000 (16:29 +0200)]
New match Reg4 type and temporary SMAC table definitions

Changes required for Bug 7236:
- Implement the match type nxm_reg_4.
- Define new source mac tables ELAN_BASE_TABLE and ELAN_SMAC_LEARNED_TABLE

Change-Id: Ib1854b86b5fd95097adcf9b90f338fff5a91659b
Signed-off-by: Olga Schukin <olga.schukin@hpe.com>
7 years agoRefactor ActionType.nx_conntrack 01/48801/6
Stephen Kitt [Tue, 29 Nov 2016 15:51:02 +0000 (16:51 +0100)]
Refactor ActionType.nx_conntrack

Change-Id: I187cbc444b996ce3574040745051e6aa7fcf89e6
Signed-off-by: Stephen Kitt <skitt@redhat.com>
7 years agoDeprecate ActionType.goto_table 00/48800/6
Stephen Kitt [Tue, 29 Nov 2016 15:13:21 +0000 (16:13 +0100)]
Deprecate ActionType.goto_table

It isn't used any more and doesn't do anything anyway.

Change-Id: I0db54f26eb9ce639f62dfa9ca8a3d3bf6b8db0fe
Signed-off-by: Stephen Kitt <skitt@redhat.com>
7 years agoRefactor push/pop MPLS/PBB/VLAN actions 87/48787/6
Stephen Kitt [Tue, 29 Nov 2016 14:16:54 +0000 (15:16 +0100)]
Refactor push/pop MPLS/PBB/VLAN actions

Change-Id: I98055271b0924c7828d1f1b96b863f8cbc96ac15
Signed-off-by: Stephen Kitt <skitt@redhat.com>
7 years agoMigrate ActionType.output to ActionOutput 82/48782/7
Stephen Kitt [Tue, 29 Nov 2016 12:35:36 +0000 (13:35 +0100)]
Migrate ActionType.output to ActionOutput

Change-Id: Ib4a469c32a5206a3cc06cb4c0895d9e57e8906ac
Signed-off-by: Stephen Kitt <skitt@redhat.com>
7 years agoMigrate ActionType.group to ActionGroup 76/48776/8
Stephen Kitt [Tue, 29 Nov 2016 11:11:55 +0000 (12:11 +0100)]
Migrate ActionType.group to ActionGroup

ActionType.group is backwards-compatible and can handle both
ActionInfo (old-style) and ActionGroup (new-style).

Change-Id: I963b80b68018aa3a827400381186695e97822cc3
Signed-off-by: Stephen Kitt <skitt@redhat.com>
7 years agoUpdated pipeline 60/48360/4
Vishal Thapar [Tue, 15 Nov 2016 11:43:47 +0000 (17:13 +0530)]
Updated pipeline

1. Fixed high level pipeline for internal tunnels
2. Created images folder for inividual file for each ascii diagram
3. Separated pipeline ascii diagrams into separate files
4. Added pipelines for L2, In/Egress ACLs and IPv6
5. Added blank file for L3, content TBD.

Change-Id: I01ae092751683bca60aeaf429f9412d32d60408f
Signed-off-by: Vishal Thapar <vishal.thapar@ericsson.com>
7 years agoMerge "OFTunnels: IFM changes"
Faseela K [Sun, 4 Dec 2016 14:16:46 +0000 (14:16 +0000)]
Merge "OFTunnels: IFM changes"

7 years agoAdd new ActionInfo implementations for reg load/move 36/48736/8
Stephen Kitt [Mon, 28 Nov 2016 12:46:21 +0000 (13:46 +0100)]
Add new ActionInfo implementations for reg load/move

This illustrates a pattern for ActionInfo implementations with strict
typing and less coupling in ActionInfo/ActionType.

Change-Id: I5bcbe039ecc5d871ef9b1ee11c88822231e92ef7
Signed-off-by: Stephen Kitt <skitt@redhat.com>
7 years agoOFTunnels: IFM changes 94/48694/7
Vishal Thapar [Thu, 24 Nov 2016 07:58:46 +0000 (13:28 +0530)]
OFTunnels: IFM changes

1. Create/Delete of tunnel port in OVSDB when tunnel interfaces are
created/deleted.
2. Additional egressActions for tunnel interfaces if source/remote IP are
set to flow.
3. Aditional matches in Ingress flow if source/remote IP are set to flow.
4. Use batching when adding child interfaces for better performance.
5. Use cache to avoid DS reads for BridgeEntry and BridgeRefEntry.

Change-Id: I4da995eba1ef3b16b05f82c4930ca4a7dcd3e58a
Signed-off-by: Vishal Thapar <vishal.thapar@ericsson.com>
7 years agoMerge "Fcaps: changing alarm text parameter to be same while raising and clearing"
Faseela K [Sat, 3 Dec 2016 00:51:40 +0000 (00:51 +0000)]
Merge "Fcaps: changing alarm text parameter to be same while raising and clearing"

7 years agoBug 7220 - OVS egress table (220) contains stale rules that send the packet to the... 22/48922/2
Faseela K [Fri, 2 Dec 2016 12:48:12 +0000 (18:18 +0530)]
Bug 7220 - OVS egress table (220) contains stale rules that send the packet to the wrong port

Description : There was a race condition in the way unbind of default egress service was done,
when the interface-state deletion happens parallelly.
Unbind Service needs interface-state information which is deleted already, Similarly stateUnbindHandler
needs service-info which is already deleted by unbind-service(), and both the threads won't
proceed with table 220 flow deletion. This is a corner case, and won't be seen always, however
it has become more visible in recent csit runs. Moving the unbindDefaultEgressDispatcherService()
logic to the EgressServiceStateUnbindHelper to avoid the race, so that it gets called once all
default 220 flows are removed from the switch.

Change-Id: Iec20a3e09064064bd5564dac2e65500c73dfad50
Signed-off-by: Faseela K <faseela.k@ericsson.com>
7 years agoMerge "Gateway mac table should have unique MAC address for vhu hosts other than...
Faseela K [Fri, 2 Dec 2016 12:31:12 +0000 (12:31 +0000)]
Merge "Gateway mac table should have unique MAC address for vhu hosts other than 00:00:00:00:00:00"

7 years agoMerge "ActionInfoBuilderTest to illustrate ActionInfoBuilder requirements"
Michael Vorburger [Fri, 2 Dec 2016 10:14:57 +0000 (10:14 +0000)]
Merge "ActionInfoBuilderTest to illustrate ActionInfoBuilder requirements"

7 years agoSingleTransactionDataBroker should not be a @Singleton with @Inject 84/48884/1
Michael Vorburger [Thu, 1 Dec 2016 13:42:30 +0000 (14:42 +0100)]
SingleTransactionDataBroker should not be a @Singleton with @Inject

because this causes issues with the blueprint-maven-plugin, see full
analysis on today's email thread "Genius CSIT Unstable" on genius-dev,
and summary just added to
https://wiki.opendaylight.org/view/BestPractices/DI_Guidelines#Nota_Bene

instead, anyone use of it can just do new
SingleTransactionDataBroker(dataBroker), which this change does for the
2 places I had used it so far.

Change-Id: Ic08184d5c4a0147ec134cedaae641dd394fd69d0
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoBug 7270 Duplicate remote Mcast mac entry in TOR . 62/48862/1
Akash [Thu, 1 Dec 2016 10:24:47 +0000 (15:54 +0530)]
Bug 7270 Duplicate remote Mcast mac entry in TOR .

As the case was not taken care the mac entries in remote Ucast mac
table for TOR device were populated with duplicate mac entries , once
with lower case and once with upper case.

Change-Id: I52b1863ebb779f22705fafb14af6289cc1234317
Signed-off-by: Akash <a.k.sahu@ericsson.com>
7 years agoFcaps: changing alarm text parameter to be same while raising and clearing 19/48819/5
gowri [Wed, 30 Nov 2016 08:34:49 +0000 (14:04 +0530)]
Fcaps: changing alarm text parameter to be same while raising and clearing

Change-Id: I941aef182be4d3a1488399d8ea1eaa7432c707ff
Signed-off-by: gowri <gowri.r@ericsson.com>
7 years agoGateway mac table should have unique MAC address for vhu hosts other than 00:00:00... 31/48731/4
Faseela K [Mon, 28 Nov 2016 10:43:54 +0000 (16:13 +0530)]
Gateway mac table should have unique MAC address for vhu hosts other than 00:00:00:00:00:00

Description : With OVS DPDK ports, the port physical address is always coming as "00:00:00:00:00:00".
Because of this interface-manager stores the phyaddress of all interfaces on top of this port
as "00:00:00:00:00:00", there by responding to the ARP requests with this MAC address for
netvirt L3VPN usecases.
For address this issue, whenever an openflow port added notification comes with hardware address "00:00:00:00:00:00",
a mac address is generated for the same using its port-number and de:ad:be:ef prefix,
and this generated mac-address will be used in all subsequent processing.

Change-Id: I305288e44d12993f8f1bad9932c19ca86fa6afca
Signed-off-by: Faseela K <faseela.k@ericsson.com>
7 years agoActionInfoBuilderTest to illustrate ActionInfoBuilder requirements 32/48832/2
Michael Vorburger [Wed, 30 Nov 2016 13:58:36 +0000 (14:58 +0100)]
ActionInfoBuilderTest to illustrate ActionInfoBuilder requirements

This should be useful as we build additional ActionInfo types, such as
the work in https://git.opendaylight.org/gerrit/#/c/48736/

Change-Id: I4aeb925784c285bea4cbf1cf65c1c057f6ae4843
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoMerge "idmanager Checkstyle now enforced, incl. much work on exception handling"
Faseela K [Wed, 30 Nov 2016 16:24:56 +0000 (16:24 +0000)]
Merge "idmanager Checkstyle now enforced, incl. much work on exception handling"

7 years agoMerge "BUG 6952: DPN can't be added in multiple TZ"
Vishal Thapar [Wed, 30 Nov 2016 09:53:12 +0000 (09:53 +0000)]
Merge "BUG 6952: DPN can't be added in multiple TZ"

7 years agoMerge "AsyncEventsWaiter with TestableDataTreeChangeListener"
Faseela K [Wed, 30 Nov 2016 08:02:36 +0000 (08:02 +0000)]
Merge "AsyncEventsWaiter with TestableDataTreeChangeListener"

7 years agoBUG 6952: DPN can't be added in multiple TZ 16/48716/2
Vishal Thapar [Fri, 25 Nov 2016 21:56:54 +0000 (03:26 +0530)]
BUG 6952: DPN can't be added in multiple TZ

In dpn-endpoints, replace transport-zone with a list tz-membership which
will track transport zones to which a DPN belongs.

Some more changes needed for delete-use case. They will come in separate
patch.

Change-Id: I77848ed142bbb0cb6e8f8489b3a28a0bd525663c
Signed-off-by: Vishal Thapar <vishal.thapar@ericsson.com>
7 years agoMerge "Fix an error on table ranges."
Faseela K [Tue, 29 Nov 2016 16:24:27 +0000 (16:24 +0000)]
Merge "Fix an error on table ranges."

7 years agoFix an error on table ranges. 93/48793/1
David Suarez [Tue, 29 Nov 2016 15:22:10 +0000 (16:22 +0100)]
Fix an error on table ranges.

-

Change-Id: I7e2ea96eaee07d49bcfcb9063fbad93925c40d4e
Signed-off-by: David Suarez <david.suarez.fuentes@ericsson.com>
7 years agoidmanager Checkstyle now enforced, incl. much work on exception handling 84/48684/8
Michael Vorburger [Thu, 24 Nov 2016 23:02:36 +0000 (00:02 +0100)]
idmanager Checkstyle now enforced, incl. much work on exception handling

Change-Id: I96d9494254bd2f782bd37d02e1cee59fbe08ca16
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoEnhancing the junits for interface-manager 78/48478/7
Faseela K [Mon, 21 Nov 2016 05:33:47 +0000 (11:03 +0530)]
Enhancing the junits for interface-manager

Change-Id: I5fd00e4e386482b3f65fcd686070056a934b14c5
Signed-off-by: Faseela K <faseela.k@ericsson.com>
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoMerge "@deprecate DataStoreCache - it's more of a joke than a real cache IMHO"
Faseela K [Mon, 28 Nov 2016 05:39:56 +0000 (05:39 +0000)]
Merge "@deprecate DataStoreCache - it's more of a joke than a real cache IMHO"

7 years agoMerge "idmanager remove usage of @deprecated CacheUtil"
Faseela K [Sat, 26 Nov 2016 12:16:48 +0000 (12:16 +0000)]
Merge "idmanager remove usage of @deprecated CacheUtil"

7 years agoidmanager remove usage of @deprecated CacheUtil 82/48682/3
Michael Vorburger [Thu, 24 Nov 2016 20:08:38 +0000 (21:08 +0100)]
idmanager remove usage of @deprecated CacheUtil

Specifically, this also solves a problem in the
InterfaceManagerConfigurationTest (InterfaceManagerTest) related to
stale content surviving between tests in this cache, due to static (that
problem could have been "resolved" via CacheUtil.destroyCache() as well,
but this is much cleaner).

Because the IdManagerCacheCli cannot use the CacheUtil with static
methods anymore now, we introduce a simple IdManagerMonitor interface,
have IdManager implement that to expose the ConcurrentMap<String,
IdLocalPool> localPool, just as Map<String,String> so as to not leak the
IdLocalPool out of the implementation.

With a little bit of blueprint XML wiring magic, the IdManagerCacheCli
can lookup that IdManagerMonitor from the OSGi service registry, and do
the same as it did before.

FTR: I've manually non-regression tested this, and verified that
genius/karaf with this change still comes up without errors, and that
the "idmanager:show" command does not blow up when used.

Change-Id: I47b92ece98fae0bce4390472c18f6833b43e7843
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years ago@deprecate CacheUtil 80/48680/3
Michael Vorburger [Thu, 24 Nov 2016 19:13:36 +0000 (20:13 +0100)]
@deprecate CacheUtil

For an example of how to simply replace this fairly useless utility
class with a ConcurrentHashMap in a project, please see the upcoming
next Gerrit doing that in idmanager.

Change-Id: I2c512aafba34a8eef83d3935c12f7048b96af471
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoDe-static-ify idmanager IdUtils 74/48674/3
Michael Vorburger [Thu, 24 Nov 2016 16:47:03 +0000 (17:47 +0100)]
De-static-ify idmanager IdUtils

As I've said before: "static is evil" ... ;-)

This isn't theory - but practice; in something like the new
InterfaceManagerTest (InterfaceManagerConfigurationTest), this static
leads to non-isolated tests - because it "leaks" state.

Change-Id: I2064ca59eba4fda5ec24e9c4c2e0dfae79a56d58
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoMerge "Bug 7230: tun_id from vxlan tunnel is incorrectly stored into gre key"
Faseela K [Fri, 25 Nov 2016 10:36:13 +0000 (10:36 +0000)]
Merge "Bug 7230: tun_id from vxlan tunnel is incorrectly stored into gre key"

7 years agoBug 7230: tun_id from vxlan tunnel is incorrectly stored into gre key 63/48663/5
Maria Merida [Thu, 24 Nov 2016 13:19:37 +0000 (14:19 +0100)]
Bug 7230: tun_id from vxlan tunnel is incorrectly stored into gre key

"tun_id" should be "cleaned" (or set to 0) when the output is a L2GRE interface.
This way we prevent a wrong "tun_id" value from being copied into the gre key field of the packet.
Right now, "tun_id" value from vxlan tunnels between OVS is not cleaned before output the packet to the GRE interface. As a result, a wrong key is set in the output packet.
More information: https://bugs.opendaylight.org/show_bug.cgi?id=7230

This fix needs to go to stable/boron SR2

Change-Id: I946b135320b4a54ba1e77f91a506d72b475d4927
Signed-off-by: Maria Merida <maria.merida@ericsson.com>
7 years agoMerge "minor: idmanager UpdateIdEntryJob else block, for clarity when reading"
Faseela K [Fri, 25 Nov 2016 05:19:58 +0000 (05:19 +0000)]
Merge "minor: idmanager UpdateIdEntryJob else block, for clarity when reading"

7 years agoFix buggy bad exception root cause logging in idmanager 73/48673/3
Michael Vorburger [Thu, 24 Nov 2016 16:04:24 +0000 (17:04 +0100)]
Fix buggy bad exception root cause logging in idmanager

Change-Id: I99fdc35e5e621f1f3149e9b5aff0ebc152700785
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agominor: idmanager UpdateIdEntryJob else block, for clarity when reading 76/48676/2
Michael Vorburger [Thu, 24 Nov 2016 18:01:26 +0000 (19:01 +0100)]
minor: idmanager UpdateIdEntryJob else block, for clarity when reading

Change-Id: I1d093bfd157134a9b9b68dc1922d2a62b0ac57cf
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years ago@deprecate DataStoreCache - it's more of a joke than a real cache IMHO 79/48679/2
Michael Vorburger [Thu, 24 Nov 2016 18:46:01 +0000 (19:46 +0100)]
@deprecate DataStoreCache - it's more of a joke than a real cache IMHO

I see no evidence in the DataStoreCache code that it is "feeded by a
clustered data change listener" (quote from the original JavaDoc).

Change-Id: I8f51bde2e72a48f80e0dfcb9f60306ae2cc85305
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoidmanager Checkstyle enforcement preparation: Organize Imports & Tabs 75/48675/2
Michael Vorburger [Thu, 24 Nov 2016 17:17:36 +0000 (18:17 +0100)]
idmanager Checkstyle enforcement preparation: Organize Imports & Tabs

incl. some other trivial automatic code formattings stuff, all low risk

incl. line length wrapping

Does not yet enforce Checkstyle, because there are a few things left to
tidy up, including reviewing Exception handling; that will come in
another future Gerrit, once this is merged.

Change-Id: Ic7d9f008ff9958b57a36cb8f0b8e1fddd3990c6e
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agominor: Some very minor first clean up in IdManager 70/48670/3
Michael Vorburger [Thu, 24 Nov 2016 15:56:05 +0000 (16:56 +0100)]
minor: Some very minor first clean up in IdManager

Change-Id: I693b60d89fd26ae36f5268fb137e2b945009801b
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoMerge "Add utility apis"
Faseela K [Thu, 24 Nov 2016 12:47:41 +0000 (12:47 +0000)]
Merge "Add utility apis"

7 years agoAsyncEventsWaiter with TestableDataTreeChangeListener 77/48077/10
Michael Vorburger [Mon, 7 Nov 2016 17:31:29 +0000 (18:31 +0100)]
AsyncEventsWaiter with TestableDataTreeChangeListener

for AsyncDataTreeChangeListenerBase etc.

see https://git.opendaylight.org/gerrit/#/c/48078/ for a usage example

Change-Id: I400acba12ec9e2bdd61a6c680fc925bde60b6686
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoAdd utility apis 39/48039/5
Sridhar Gaddam [Mon, 7 Nov 2016 07:39:42 +0000 (13:09 +0530)]
Add utility apis

This patch implements the following utility apis.
1. API to check if the supplied ipAddress is IPv4 Address or IPv6
2. API that returns the corresponding ipPrefix based on the ipAddress

Change-Id: I975e23763d45ceef3598ac35970767555196efe7
Signed-off-by: Sridhar Gaddam <sgaddam@redhat.com>
7 years agoMerge "OFTunnels: Convert ITM config to IFM"
Faseela K [Thu, 24 Nov 2016 10:46:13 +0000 (10:46 +0000)]
Merge "OFTunnels: Convert ITM config to IFM"

7 years agoMerge ".gitignore .fbExcludeFilterFile"
Michael Vorburger [Thu, 24 Nov 2016 10:43:13 +0000 (10:43 +0000)]
Merge ".gitignore .fbExcludeFilterFile"

7 years agoTighten DefaultBatchHandler properties 31/48331/2
Stephen Kitt [Mon, 14 Nov 2016 17:37:31 +0000 (18:37 +0100)]
Tighten DefaultBatchHandler properties

* dataBroker, batchSize and batchInterval shouldn't be static.
* None of the properties should be public.

Change-Id: Ibd0ee0e35566824ea4aedfaa5766d5de8a830f95
Signed-off-by: Stephen Kitt <skitt@redhat.com>
7 years agoMerge "SingleTransactionDataBroker"
David Suarez [Wed, 23 Nov 2016 23:46:26 +0000 (23:46 +0000)]
Merge "SingleTransactionDataBroker"

7 years ago.gitignore .fbExcludeFilterFile 25/48625/1
Michael Vorburger [Wed, 23 Nov 2016 13:01:00 +0000 (14:01 +0100)]
.gitignore .fbExcludeFilterFile

Change-Id: I92d9aef533fd001e243efa98aad8c34696efb80a
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoOFTunnels: Convert ITM config to IFM 12/48612/1
Vishal Thapar [Wed, 23 Nov 2016 10:17:53 +0000 (15:47 +0530)]
OFTunnels: Convert ITM config to IFM

This change uses OFTunnels configuration in ITM to create tunnel
interfaces with remote-ip-flow option set.

Change-Id: I74f6b494d23137516608005a9b3cb96184031eb8
Signed-off-by: Vishal Thapar <vishal.thapar@ericsson.com>
7 years agoOFTunnels: Update spec and itm-state.yang 08/48608/1
Vishal Thapar [Wed, 23 Nov 2016 08:18:42 +0000 (13:48 +0530)]
OFTunnels: Update spec and itm-state.yang

One more addition made to itm-state.yang to optimize passing of of-tunnel
option from TransportZone configuration to Interface configuration. This
change eliminates need to do a DS read.

Spec also updated to reflect this addition.

Change-Id: I9396836294bf9d01573acbeddb6a35134610e63a
Signed-off-by: Vishal Thapar <vishal.thapar@ericsson.com>
7 years agoSingleTransactionDataBroker 52/48452/4
Michael Vorburger [Thu, 17 Nov 2016 18:30:28 +0000 (19:30 +0100)]
SingleTransactionDataBroker

Change-Id: I9a88fe7672235057329968481b8f3438685b0859
Signed-off-by: Michael Vorburger <vorburger@redhat.com>
7 years agoMerge "Cleanup: keySet() never returns null"
David Suarez [Mon, 21 Nov 2016 07:48:00 +0000 (07:48 +0000)]
Merge "Cleanup: keySet() never returns null"

7 years agoMerge "Cleanup Optional uses"
David Suarez [Mon, 21 Nov 2016 07:47:21 +0000 (07:47 +0000)]
Merge "Cleanup Optional uses"

7 years agoMerge "Cleanup: use lambdas and method references"
David Suarez [Mon, 21 Nov 2016 07:47:13 +0000 (07:47 +0000)]
Merge "Cleanup: use lambdas and method references"

7 years agoMerge "Addition of constants for ARP Responder"
Faseela K [Mon, 21 Nov 2016 07:30:21 +0000 (07:30 +0000)]
Merge "Addition of constants for ARP Responder"

7 years agoMerge "bug 7205 l2gw itm mesh is not getting built"
Faseela K [Mon, 21 Nov 2016 06:51:16 +0000 (06:51 +0000)]
Merge "bug 7205 l2gw itm mesh is not getting built"

7 years agoIgnoring new interfacemanager tests added as component style tests. 90/48490/1
Faseela K [Fri, 18 Nov 2016 16:40:36 +0000 (22:10 +0530)]
Ignoring new interfacemanager tests added as component style tests.

Description : Unit test failures introduced after this.
Will be revisited

Change-Id: I2da3c03163f9e858aa236710d9982e5657f5da8c
Signed-off-by: Faseela K <faseela.k@ericsson.com>
7 years agobug 7205 l2gw itm mesh is not getting built 75/48475/1
K.V Suneelu Verma [Fri, 18 Nov 2016 12:37:33 +0000 (18:07 +0530)]
bug 7205 l2gw itm mesh is not getting built

If there is default transport zone which is created as part of controller
start rpc call to mesh l2gw device to itm is failing.

Change-Id: I87b941976b42544e53a567cc8fb9264e9c6ba075
Signed-off-by: K.V Suneelu Verma <k.v.suneelu.verma@ericsson.com>
7 years agoMerge "Cleanup: remove unnecessary boxing/unboxing"
David Suarez [Fri, 18 Nov 2016 12:12:05 +0000 (12:12 +0000)]
Merge "Cleanup: remove unnecessary boxing/unboxing"