[CONTROLLER-643] OpenDaylight "fails to fail" on mounting device which doens't provide schemas if there are other schemas in cache Created: 23/Jul/14 Updated: 25/Jul/23 Resolved: 25/Mar/15 |
|
| Status: | Resolved |
| Project: | controller |
| Component/s: | netconf |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | ||
| Reporter: | Giles Heron | Assignee: | Unassigned |
| Resolution: | Won't Do | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Operating System: Linux |
||
| External issue ID: | 1408 |
| Description |
|
There's a test in NetconfDeviceSchemaProviderFactory.java which (I believe) is intended to check if a device has not provided any schemas: Preconditions.checkState(sourceContext.getValidSources().isEmpty() == false, however this check only fails if the ~/cache/schema directory is completely empty. If there are files in the directory which were provided by another device then the check "fails to fail" and the device is mounted despite failing to provide any sources. Is this intentional? |
| Comments |
| Comment by Tony Tkacik [ 24/Jul/14 ] |
|
Hi Giles, If device does not implement ietf-netconf-monitoring, we still looks up for |
| Comment by Giles Heron [ 24/Jul/14 ] |
|
ah, ok. so we just check for any one of the models supported being in the cache? did we consider checking for all of them? or just not bothering to check? I mean if one module is there how much better is that than none? |
| Comment by Tony Tkacik [ 24/Jul/14 ] |
|
So current behaviour is to load as much as possible models, |
| Comment by Tony Tkacik [ 16/Sep/14 ] |
|
The behaviour of netconf connector is to try to get sources from cache / other devices if they have same revision of same model. |
| Comment by Maros Marsalek [ 19/Mar/15 ] |
|
Hi Giles, So this is expected behavior as sal-netconf-connector tries to build the SchemaContext from the biggest possible subset of sources. Netconf-connector logs the issues with resolving sources and recently, it also writes the problematic sources into the datastore and they are available using restconf in topology model e.g. http://localhost:8181/restconf/operational/network-topology:network-topology/topology/topology-netconf/ So I would close this issue as expected behavior, what do you say ? |
| Comment by Maros Marsalek [ 25/Mar/15 ] |
|
This is expected behaviour, netconf node in the netconf topology now lists all the schemas that could not be used. |