[MDSAL-134] java.lang.IllegalStateException: Unsupported DOM Modification type UNMODIFIED Created: 08/Mar/16 Updated: 04/Aug/18 Resolved: 04/Aug/18 |
|
| Status: | Resolved |
| Project: | mdsal |
| Component/s: | Binding runtime |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | ||
| Reporter: | Moiz Raja | Assignee: | Unassigned |
| Resolution: | Cannot Reproduce | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Operating System: All |
||
| External issue ID: | 5468 |
| Description |
|
I am seeing this exception happen in Lithium. This happens when I getModifiedChildren on a DataObjectModification and call getModificationType on one of the modified children. Why would this happen? 2016-03-05 00:15:12.844 GMT+00:00 [opendaylight-cluster-data-notification-dispatcher-54] ERROR o.o.c.c.d.DataTreeChangeListenerActor - Error notifying listener org.opendaylight.controller.md.sal.binding.impl.BindingClusteredDOMDataTreeChangeListenerAdapter@79048d7e |
| Comments |
| Comment by Tony Tkacik [ 08/Mar/16 ] |
|
Effectivelly it means BindingDataTreeChangeListener received from clustering DataTreeCandidateNode with UNMODIFIED which means listener was triggered even if data were not modified. |
| Comment by Tom Pantelis [ 08/Mar/16 ] |
|
I assume that can happen if a client wrote the same data? |
| Comment by Moiz Raja [ 08/Mar/16 ] |
|
I think the bug is not in the generating of the notification but in the implementation of LazyDataObjectModification#getModifiedChildren(). It should not return a child which has the modificationType UNMODIFIED. This is happening in my specific implementation when I listen at a top-level node and when I get a data tree change I navigate the modification tree to figure out whether a specific subtree change happened. |
| Comment by Robert Varga [ 08/Mar/16 ] |
|
DTCL does not perform a comparison of the actual data, so an overwrite with the same data still shows up as WRITE. |
| Comment by Robert Varga [ 27/Oct/16 ] |
|
Does this still happen with Boron/Carbon? Do we have a unit test? |
| Comment by Tom Pantelis [ 27/Oct/16 ] |
|
I haven't seen this exception. It doesn't appear there's any test case available at this point. |
| Comment by Harinath Mallepally [ 01/Mar/17 ] |
|
I am using BE SR3 and I see this exception |
| Comment by Srini Seetharaman [ 03/Apr/17 ] |
|
It happens with Boron-SR2 too. |
| Comment by Robert Varga [ 30/Jan/18 ] |
|
We will need at least some data on what the model/DS operation triggering this is, otherwise we can only guess as to what is going on. |