[OVSDB-236] southbound fails to delete OvsdbNode from config and operational Created: 10/Dec/15  Updated: 19/Oct/17  Resolved: 18/Dec/15

Status: Resolved
Project: ovsdb
Component/s: Southbound.Open_vSwitch
Affects Version/s: unspecified
Fix Version/s: None

Type: Bug
Reporter: Sam Hague Assignee: Unassigned
Resolution: Duplicate Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Operating System: All
Platform: All

Attachments: Text File 4756.karaf.log     Text File karaf.log    
Issue Links:
Duplicate
duplicates YANGTOOLS-559 Delete transactions fail in MinMaxEle... Resolved
External issue ID: 4756

 Description   

Adding a OvsdbNode via RESTCONF and then deleting it via RESTCONF fails to remove the OvsdbNode from config and operational.

I see exceptions like below. I think maybe the ownership has changed or been removed for this ODL and the pending mdsal updates fail because there is no owner. Some longer time later the config and operational do cleanup. karaf log is attached.

2015-12-10 16:16:03,081 | WARN | lt-dispatcher-17 | SimpleShardDataTreeCohort | 163 - org.opendaylight.controller.sal-distributed-datastore - 1.3.0.SNAPSHOT | Unexpected failure in validation phase
com.google.common.base.VerifyException: Total child count is -1 (from NodeModification [identifier=(urn:TBD:params:xml:ns:yang:network-topology?revision=2013-10-21)node,

Steps:

1. start ovsdb to pssive listen: sudo ovs-vsctl set-manager ptcp:6640
2. RESTCONF to add node [add]
3. RSTCONF to delete node [del]
4. ODL disconnects the ovsdb node.
5. exceptions
6. GET to see that both config and operation still exist. [get]
7. Some time later the mdsal does cleanup such that the GET's return nothing.

[add]
http://localhost:8080/restconf/config/network-topology:network-topology/topology/ovsdb:1/node/ovsdb:%2F%2F192.168.120.31:6640
{
"network-topology:node": [
{
"node-id": "ovsdb://192.168.120.31:6640",
"connection-info":

{ "ovsdb:remote-port": 6640, "ovsdb:remote-ip": "192.168.120.31" }

}
]
}

[del]
http://localhost:8080/restconf/config/network-topology:network-topology/topology/ovsdb:1/node/ovsdb:%2F%2F192.168.120.31:6640

[get]
http://localhost:8080/restconf/config/network-topology:network-topology/

 Comments   
Comment by Sam Hague [ 10/Dec/15 ]

Attachment karaf.log has been added with description: karaf log showing exceptions

Comment by Jamo Luhrsen [ 16/Dec/15 ]

a different repro:

1. install odl-ovsdb-southbound-impl-rest
2. sudo ovs-vsctl set-manager tcp:${ODL_IP}:6640
3. sudo ovs-vsctl del-manager

doing the above with a distro built on 12/15 (master) [0] showed these
symptoms. karaf log attached.

[0] distribution-karaf-0.4.0-20151215.221738-2802.zip

Comment by Jamo Luhrsen [ 16/Dec/15 ]

Attachment 4756.karaf.log has been added with description: karaf log

Comment by Jamo Luhrsen [ 16/Dec/15 ]

Also just hit this by having an ovs actively trying to connect while brining
up karaf and installing ovsdb-southbound-impl-rest. It did connect, but when
disconnecting, I saw the Exception in karaf.log

Comment by Sam Hague [ 17/Dec/15 ]

Good point to add that if the odl-ovsdb-southbound-impl-ui feature is used rather than odl-ovsdb-southbound-rest then the error does not occur. The -ui feature starts with -rest and adds odl-mdsal-apidocs and odl-mdsal-xsql.

Comment by Sam Hague [ 18/Dec/15 ]

(In reply to Sam Hague from comment #4)
> Good point to add that if the odl-ovsdb-southbound-impl-ui feature is used
> rather than odl-ovsdb-southbound-rest then the error does not occur. The -ui
> feature starts with -rest and adds odl-mdsal-apidocs and odl-mdsal-xsql.

Should also note that this could be mere coincidence that the -ui feature doesn't show the problem. The issue seems to be timing related so the -ui could just be masking the issue.

Generated at Wed Feb 07 20:35:51 UTC 2024 using Jira 8.20.10#820010-sha1:ace47f9899e9ee25d7157d59aa17ab06aee30d3d.