Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: spelling and grammar

In order to integrate legacy devices, there is a common approach to implement an adapter or mediator software, This software usually runs in a data-center and provides northbound the target interface and data modules and consumes on the southbound side legacy protocols and data models.

In cases where the legacy management protocol bases is based on UDP, there is a good chance that important notifications are not forward forwarded to ONAP, because the UDP packet gets lost. 

It happens for sure, in case of in-band management , and for alarms indicating traffic errors. 

Due to the fact that such connection loss can not be detected by SDN-R, there should be a polling mechanism implemented to request all air-interface problem list every 120s - this is possible for ONF-CoreModel 1.2, as the expected deployed NetConf devices/mediators are quite limited. 

In the target architecture using another version of the ONF-CoreModel or other fault management modules (e.g. ietf-alarms) - the workaround would not apply, because the OAM connection would be detected by SDN-R and the normal alarm synchronization process would take place. 


Jira
serverONAP JIRA
columnskey,summary,priority,status
maximumIssues20
jqlQuerykey = SDNC-1070
serverId425b2b0a-557c-3c0c-b515-579789cceedb

draw.io Diagram
bordertrue
diagramNameworkaround-for-udp-notifications
simpleViewerfalse
width
linksauto
tbstyletop
lboxtrue
diagramWidth862
revision1