vCPE Use Case - Customer Service Instantiation - 171103.pdf
11/9/2017
- vCpeResCust custom workflow:
- Brian made changes on the SDNC side. Now SDNC can pass a list of parameters to SO for vG assign call. Then SO passes those parameters to HEAT to instantiate vG.
- Jim fixed the DoDeleteVfModule flow to use generic-resource-api and construct the corresponding request body when performing SDNC deactivate call. Note that the flow checks the configuration variable sdncversion to determine what request body to construct. This is something not fully understood by the team.
- Delete of vG succeeded.
- Jim continues to work on
.Jira server ONAP JIRA columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution serverId 425b2b0a-557c-3c0c-b515-579789cceedb key SO-325 Eric has modified vGMUX to add a workaround, see below:
Code Block title vGMUX set sourceName collapse true There is a vgmux image in the ONAP-vCPE project space called: vgmux2-base-ubuntu-16-04 This one has the ability to configure the sourceName in the VES event to something different than the default value (which is the vnf-id present in the vm’s openstack metadata). Some documentation: Configuring the VES mode - via REST This will set the ‘demo’ mode and packet loss to 40%, but does ‘not’ change the sourceName: curl -i -H "Content-Type:application/json" --data '{"mode":{"working-mode":"demo","base-packet-loss":40,"source-name":""}}' -X POST -u admin:admin http://127.0.0.1:8183/restconf/config/vesagent:vesagent Delete the config in order to change it via REST: curl -i -H "Content-Type:application/json" -X DELETE -u admin:admin http://127.0.0.1:8183/restconf/config/vesagent:vesagent/mode curl -i -H "Content-Type:application/json" --data '{"mode":{"working-mode":"demo","base-packet-loss":88,"source-name":"testing-123-ABC"}}' -X POST -u admin:admin http://127.0.0.1:8183/restconf/config/vesagent:vesagent Configuring the VES mode - via CLI QUERY: # vppctl show ves mode Mode Base Packet Loss Rate Source Name Demo 88.0% testing-123-ABC SET: vppctl set ves mode demo base 77 source hello-there This sets the sourceName to “hello-there” Leave off the 'source <name>' arguments to set back to default (i.e. vnf-id from openstack metadata) Sample event with an overwritten sourceName: { "event": { "commonEventHeader": { "domain": "measurementsForVfScaling", "eventId": "Generic_traffic", "eventName": "Measurement_vGMUX", "eventType": "HTTP request rate", "lastEpochMicrosec": 1510347222243201, "priority": "Normal", "reportingEntityId": "No UUID available", "reportingEntityName": "zdcpe1cpe01mux01", "sequence": 23, "sourceId": "vCPE_Infrastructure_vGMUX_demo_app", "sourceName": "testing-123-ABC", "startEpochMicrosec": 1510347212243201, "version": 1.2 }, "measurementsForVfScalingFields": { "additionalMeasurements": [ { "arrayOfFields": [ { "name": "Packet-Loss-Rate", "value": "88.0" } ], "name": "ONAP-DCAE" } ], "cpuUsageArray": [ { "cpuIdentifier": "cpu1", "cpuIdle": 100.0, "cpuUsageSystem": 0.0, "cpuUsageUser": 0.0, "percentUsage": 0.0 } ], "measurementInterval": 10, "measurementsForVfScalingVersion": 2.1, "requestRate": 2567, "vNicUsageArray": [ { "receivedOctetsDelta": 0.0, "receivedTotalPacketsDelta": 0.0, "transmittedOctetsDelta": 0.0, "transmittedTotalPacketsDelta": 0.0, "vNicIdentifier": "eth0", "valuesAreSuspect": "true" } ] } } }
11/9/2017
- vCpeResCust custom workflow:
- Closed loop: manually send VES event to DCAE, APPC successfully restarted VNF via two options: directly talking to Openstack, calling MultiCloud API.
- To restart through Openstack, we need the following vserver info in AAI. This is added by heat bridge. Note that if identity-url is missing then APPC will use the default one from its properties.
Code Block "vserver-selflink": "http://10.12.25.2:8774/v2.1/466979b815b5415ba14ada713e6e1846/servers/9e17027d-c09d-45c4-9a7a-84f50a2246fd",
In the appc container, /opt/openecomp/appc/data/properties/appc.properties should have either of the following lines.
Code Block provider1.identity=http://10.12.25.2:5000/v3 #this is for openstack provider1.identity=http://10.0.14.1:9005/api/multicloud-titanium_cloud/v0/pod25_RegionOne/identity/v2.0 #this is for multicloud
To restart through MultiCloud, the following ESR should be added, note that the identity-url and the vserver self link is pointing to MultiCloud. Note that the vserver selflink has a different IP and port. APPC picks up the user name and password from ESR and use them to access MultiCloud API.
Code Block title ESR and vserver data from AAI collapse true https://aai1:8443/aai/v11/cloud-infrastructure/cloud-regions/cloud-region/CloudOwner/RegionOne?depth=all { "cloud-owner": "CloudOwner", "cloud-region-id": "RegionOne", "cloud-type": "SharedNode", "owner-defined-type": "OwnerType", "cloud-region-version": "v1", "identity-url": "http://10.0.14.1:9005/api/multicloud-titanium_cloud/v0/pod25_RegionOne/identity/v3", "cloud-zone": "CloudZone", "sriov-automation": false, "resource-version": "1510239649440", "tenants": { "tenant": [ { "tenant-id": "466979b815b5415ba14ada713e6e1846", "tenant-name": "Integration", "resource-version": "1509709609665", "vservers": { "vserver": [ { "vserver-id": "9e17027d-c09d-45c4-9a7a-84f50a2246fd", "vserver-name": "zdcpe1cpe01dns01_110306", "vserver-name2": "zdcpe1cpe01dns01_110306", "prov-status": "ACTIVE", "vserver-selflink": "http://10.0.14.1:80/v2.1/466979b815b5415ba14ada713e6e1846/servers/9e17027d-c09d-45c4-9a7a-84f50a2246fd", "in-maint": false, "is-closed-loop-disabled": false, "resource-version": "1510183808208", .... "esr-system-info-list": { "esr-system-info": [ { "esr-system-info-id": "432ac032-e996-41f2-84ed-9c7a1766eb29", "system-name": "example-system-name-val-29070", "type": "example-type-val-85254", "vendor": "example-vendor-val-94515", "version": "example-version-val-71880", "service-url": "http://10.12.25.2:5000/v3", "user-name": "demo", "password": "onapdemo", "system-type": "VIM", "ssl-cacert": "example-ssl-cacert-val-75021", "ssl-insecure": true, "ip-address": "example-ip-address-val-44431", "port": "example-port-val-93234", "cloud-domain": "Default", "default-tenant": "Integration", "resource-version": "1510239649503" } ] } }
- To restart through Openstack, we need the following vserver info in AAI. This is added by heat bridge. Note that if identity-url is missing then APPC will use the default one from its properties.
- Tested DeleteVcpeResCust flow.
- After a few bug fixes, the flow was able to delete the vG stack. But the subsequent SDNC request returned an error:
.Jira server ONAP JIRA columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution serverId 425b2b0a-557c-3c0c-b515-579789cceedb key SDNC-184 - The flow failed to pick up allotted resources:
Jira server ONAP JIRA columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution serverId 425b2b0a-557c-3c0c-b515-579789cceedb key SO-325
- After a few bug fixes, the flow was able to delete the vG stack. But the subsequent SDNC request returned an error:
- Closed loop: manually send VES event to DCAE, APPC successfully restarted VNF via two options: directly talking to Openstack, calling MultiCloud API.
...