#ricp how to handle E2SM which relies on xapp // RIC-628 + RIC-444 #ricp

Thoralf Czichy



generally the RIC platform should not depend on E2SMs. On E2M side we have had an issue that this was not correctly implemented and it is now fixed as solution to https://jira.o-ran-sc.org/browse/RIC-444  (see the gerrit link in the JIRA issue).


As far as I understand there’s a similar issue in subscription manager visible as “e2err_RICEventTriggerDefinitionEmptyDecodeDefaultFail”. The discussion in the development team for this already started, but was on hold for the last two weeks due to ongoing holiday period. For now I also started a JIRA bug case for this: https://jira.o-ran-sc.org/browse/RIC-628 . I am not aware of a good workaround for this.






From: main@... <main@...> On Behalf Of Minhac Lee via lists.o-ran-sc.org
Sent: Wednesday, July 15, 2020 11:19 AM
To: main@...
Subject: [oran-sc] #ricp how to handle E2SM which relies on xapp


Hello, I'm Minha Lee.

I am writing to ask you about E2SM which relies on xapp.
I understood that `submgr` and `e2term` should have a routine(including E2SM files) to handle E2SM requested by xapp, is it right?

I tested RIC Platform with hello-world xapp and I got some error like below in submgr.

{"ts":1594796956060,"crit":"ERROR","id":"submgr","mdc":{"time":"2020-07-15T07:09:16"},"msg":"XAPP-SubReq: params(Src=service-ricxapp-hwxapp-rmr.ricxapp:4560 Mtype=12010 SubId=-1 Xid= Meid=meid(RanName=gnb:0712-1366-b5c67788) Paylens=64/64 Paymd5=56ee049a97464806a1349b9751ac9dbe) err(e2err(e2err_RICEventTriggerDefinitionDecodeWMOREFail)

As I thought, using different E2SM seems to be a cause of problem.
submgr decodes `sub_request` using `e2sm-gNB-NRT-v401.asn` but hello-world xapp doesn't use `e2sm-gNB-NRT-v401.asn` when building `sub_request`.

I'd like to know if RIC Platform components should have E2SMs requested by xapp or not
And I wonder what should be changed in RIC Platform when xapp with the new E2SM is added.