SO hangs for 5 minutes, when executing VF-Module orchestration

Description

I noticed, that SO-bpmn-infra hangs for about 5 minutes.
In the log attached (so-bpmn-infra), we have the following:

2020-05-07T12:22:32.281Z|01bbf662-7054-4838-8e9b-321354b545c7|o.apache.cxf.interceptor.OutgoingChainInterceptor - Interceptors contributed by binding: [org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor@74ae6753]
2020-05-07T12:22:32.283Z||org.apache.cxf.interceptor.LoggingOutInterceptor - Outbound Message
---------------------------
ID: 20
Response-Code: 202
Content-Type: application/json
Headers: {Content-Type=[application/json], Date=[Thu, 07 May 2020 12:22:32 GMT]}
Payload: {"response":"{\"requestReferences\":{\"requestId\":\"01bbf662-7054-4838-8e9b-321354b545c7\",\"instanceId\":\"51773ab3-db52-4940-af75-bea6cb369484\"}}","message":"Success","processInstanceID":"6c609b22-905d-11ea-98cd-82f3703e2d77","variables":null,"messageCode":200}
--------------------------------------
2020-05-07T12:27:46.716Z|01bbf662-7054-4838-8e9b-321354b545c7|o.o.s.b.c.l.f.FlowManipulatorListenerRunner - Running flow manipulators:
org.onap.so.bpmn.infrastructure.workflow.tasks.listeners.MultiStageSkipListener

Clearly there is a 5 minutes of time, where the SO-BPMN-INFRA is not really doing anything.
We observe this in different environments.

On the API (where You can monitor the progress), You see this as the VF-Module orchestration in IN_PROGRESS, but You cannot see any BBs started, and only after +/- 5 minutes, You start to see the execution of first BBs. All of this before the Request is sent to OpenStack in order to orchestrate a Heat Template (that happens in CreateVFModule BB, whichis executed after these 5 minutes break).

Doesn`t happen for Service or VNF orchestrations. Perhaps matter of SDNC preload assignment? Not really sure looking at the SO logs.

Please have a look at the log attached.

Environment

so images in version 1.6.0.

Attachments

1

Activity

Show:

Former user May 11, 2020 at 10:41 AM

Closing this Bug report.
I re-tested using .6.1 image, and I don`t see this 5 minutes hanging happening anymore.

Former user May 8, 2020 at 9:54 AM
Edited

I was not really testing with CDS here.
We have a simple environment, a`la carte orchestration using SO/GR API, we`re using SDNC Pre-load to customize the ENVs of associated Heat template.

Now, based on what I see in the log (but I might be wrong), this waiting period happens before the building blocks are actually started. We have actually a list of BBs prepared:

2020-05-07T12:22:31.486Z|01bbf662-7054-4838-8e9b-321354b545c7|o.o.s.b.i.workflow.tasks.WorkflowAction - List of BuildingBlocks to execute:
2020-05-07T12:22:31.487Z|01bbf662-7054-4838-8e9b-321354b545c7|o.o.s.b.i.workflow.tasks.WorkflowAction - AssignVfModuleBB
2020-05-07T12:22:31.487Z|01bbf662-7054-4838-8e9b-321354b545c7|o.o.s.b.i.workflow.tasks.WorkflowAction - CreateVfModuleBB
2020-05-07T12:22:31.487Z|01bbf662-7054-4838-8e9b-321354b545c7|o.o.s.b.i.workflow.tasks.WorkflowAction - ActivateVfModuleBB

But I am really not sure, if teh procssing for the 1st BB has actually been even started at that point, if any BPMN Task has been executed before that.
SO-Monitor would actually help here a lot...

Seshu Kumar Mudiganti May 7, 2020 at 11:11 PM

this seems strange, is cds flows working fine?
Did you notice anything waiting for the info from previous task?

Done

Details

Assignee

Reporter

Fix versions

Affects versions

Priority

Created May 7, 2020 at 1:43 PM
Updated May 28, 2020 at 2:08 AM
Resolved May 11, 2020 at 10:31 AM