Closed Loop for Network Slicing

Manual Configuration:

This configuration is required to enable the ranSlice features in SDN-R.

SDNC:

•  Create the below environment variable in the helm chart at kubernetes/sdnc/templates/statefulset.yaml (for sdnc container). 

+          - name: SDNR_NORTHBOUND
+            value: "{{ .Values.config.sdnr.enabled | default "false"}}"

• Make the SDNC chart and then deploy.
• Execute the following command to log into the sdnc container 
  1. kubectl exec -it dev-sdnc-0 bash 
• Once in the docker container, edit the properties file  /opt/onap/ccsdk/data/properties/ran-slice-api-dg.properties
  •  update the following flag to true.
    1. ransim-mounted=true
    2. configdb-deployed=false  (set to true if configDB is used)
    3. cps-deployed=true   (set to true if cps is used)
  • Update IP address for ConfigDB:
    1. configdb.url=${CONFIG_DB_URL:-http://<IPAddress>:8086/api/sdnc-config-db/v4}
  • Update IP address for CPS-TBDMT:
    1. restapi.cps.anchors.endpoint=${CPS_URL:-http://<IPAddress>:8883}/cps/api/v1/dataspaces/E2EDemo/anchors

POLICY:

If old image is used:

Add the below line (in bold) in the helm chart at kubernetes/policy/components/policy-drools-pdp/templates/statefulset.yaml. 

           command: ["sh","-c"]
           args: ["if [ -f {{ .Values.certInitializer.credsPath }}/.ci ]; then \
                   source {{ .Values.certInitializer.credsPath }}/.ci; fi;\
+                  echo 'actor.service.SO.operations.Modify\\ NSSI.path=3gppservices/v1/modify' >> /opt/app/policy/config/event-manager.properties; \
                   cp {{ .Values.certInitializer.credsPath }}/org.onap.policy.p12 ${POLICY_HOME}/etc/ssl/policy-keystore;\
                   /opt/app/policy/bin/pdpd-cl-entrypoint.sh boot"]

• Make the Policy chart and then deploy.

Deployment Prerequisite/dependencies

1. DMaaP, Policy, SO and SDNC pods should be up and running.

2. VES, PM mapper and DataFile Collector service should be running.
3. ConfigDB container should be running(if configDb is disabled in slice-analysis, CPS and CPS-TBDMT containers  should be running).

Slice Analysis MS

Edit the below configurations in slice-analysis helm chart oom/kubernetes/dcaegen2-services/components/dcae-slice-analysis-ms/values.yaml

• if config DB is used:

sliceanalysisms.configDbEnabled: true
sliceanalysisms.configDb.service: http://<IPAddress>:8086

• if CPS is used:

  sliceanalysisms.configDbEnabled: false
  sliceanalysisms.cps.url: http://<IPAddressOfCPS-TBDMT>:8088/execute/ran-network

Make the chart and deploy slice-analysis

Policies for Slice Analysis MS:

Execute these curl requests from any of the running pods.

Policy Creation:

curl -k --silent --user 'policyadmin:zb!XztG34' -X POST "https://policy-api:6969/policy/api/v1/policytypes/onap.policies.controlloop.operational.common.Drools/versions/1.0.0/policies" -H "Accept: application/json" -H "Content-Type: application/json" -d @modifynssi.json

where modifynssi.json is, 
SLA MS - Policy

To push the policy:

curl --silent -k --user 'policyadmin:zb!XztG34' -X POST "https://policy-pap:6969/policy/pap/v1/pdps/policies" -H "Accept: application/json" -H "Content-Type: application/json" -d @push_modifynssi.json

where push_modifynssi.json is,
SLA MS - Push the Policy

ConfigDB:

1. Download the configDb source files from latest-configdb.zip
2. cd configdb (after extracting the files)
3. Setup and run maridb container
   1. sudo docker run -p 3306:3306 -v $PWD/SDNC_ConfigDB_SchemaV4.sql:/docker-entrypoint-initdb.d/SDNC_ConfigDB_SchemaV4.sql --name mariadb -e MYSQL_ROOT_PASSWORD=password -d mariadb
4. Navigate to project directory
   1. cd configdb/Config-DB-API-App
5. Build the application
   1. JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64/ mvn clean install
6. Run the application
   1. JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64/ mvn spring-boot:run &

CPS and CPS-TBDMT:

CPS:

1. git clone "https://gerrit.onap.org/r/cps" 
2. cd cps/
3. mvn clean install
4. cd docker-compose/
5. docker-compose up

CPS-TBDMT:

1. git clone "https://gerrit.onap.org/r/cps/cps-tbdmt" 
2. cd cps-tbdmt/
3. mvn clean install
4. cd docker-compose/
5. update cps & ncmp ip-address in application.yml
6. docker-compose up (use the image built using the above steps in the docker-compose file)

Creating dataspace: 
curl --location --user cpsuser:cpsr0cks! -H "Accept: application/json" -H "Content-Type: application/json" \
--request POST \
http://$CPS_IP:8080/cps/api/v1/dataspaces?dataspace-name=E2EDemo

Creating schema set: 

ran-network.zip
curl --location --user cpsuser:cpsr0cks! \
--request POST \
http://$CPS_IP:8080/cps/api/v1/dataspaces/E2EDemo/schema-sets --form 'file=@"ran-network.zip"' --form 'schema-set-name="ran-network"'

Creating anchor:
curl --location --user cpsuser:cpsr0cks!  --request POST \
http://$CPS_IP:8080/cps/api/v1/dataspaces/E2EDemo/anchors?schema-set-name=ran-network \
-d anchor-name=ran-network-anchor

Uploading cps payload:

payload-ran-network.json
curl --location --user cpsuser:cpsr0cks! --request POST \
http://$CPS_IP:8080/cps/api/v1/dataspaces/E2EDemo/anchors/ran-network-anchor/nodes \
--header 'Content-Type: application/json' \
-d @payload-ran-network.json

Uploading tbdmt-templates:

curl --location --request POST \
http://$CPS_TBDMT_IP:8080/templates \
--header 'Content-Type: application/json' \
--data-raw '{"templateId": "get-nearrtric-config","model": "ran-network","requestType": "query-cps-path","xpathTemplate": "//sNSSAIList[@sNssai='\''{{sNssai}}'\'']/ancestor::NearRTRIC","includeDescendants": true,"transformParam":"NearRTRIC"}'

curl --location --request POST \
http://$CPS_TBDMT_IP:8080/templates \
--header 'Content-Type: application/json' \
--data-raw '{"templateId": "get-gnbdufunction-by-snssai","model": "ran-network","requestType": "query-cps-path","xpathTemplate": "//sNSSAIList[@sNssai='\''{{sNssai}}'\'']/ancestor::GNBDUFunction","includeDescendants": true,"transformParam":"GNBDUFunction"}'

curl --location --request POST \
http://$CPS_TBDMT_IP:8080/templates \
--header 'Content-Type: application/json' \
--data-raw '{"templateId": "get-nrcelldu-by-snssai","model": "ran-network","requestType": "query-cps-path","xpathTemplate": "//sNSSAIList[@sNssai='\''{{sNssai}}'\'']/ancestor::NearRTRIC","includeDescendants": true,"transformParam":"NearRTRIC"}'

curl --location --request POST\
http://$CPS_TBDMT_IP:8080/templates \
--header 'Content-Type: application/json'\
--data-raw '{
"templateId": "get-plmnid","model": "ran-inventory","requestType": "query-cps-path","xpathTemplate": 
"//sliceProfilesList[@sliceProfileId='\''{{sliceProfileId}}'\'']","includeDescendants": true,"transformParam":"sliceProfilesList,pLMNIdList"}'

curl --location --request POST\
http://$CPS_TBDMT_IP:8080/templates\
--header 'Content-Type: application/json'\
--data-raw '{"templateId": "patch-configData","model": "ran-network","requestType": "put","xpathTemplate":
"/ran-network/NearRTRIC[@idNearRTRIC='\''{{idNearRTRIC}}'\'']/attributes/pLMNInfoList[@mcc='\''{{mcc}}'\'' and @mnc='\''{{mnc}}'\'']/sNSSAIList[@sNssai='\''{{sNssai}}'\'']","includeDescendants": true}'

RANSIM and Honeycomb setup:

RANSIM :

Follow the steps below to setup the Ransim environment:

1. Clone the repo from : git clone https://gerrit.onap.org/r/integration/simulators/ran-simulator

2. Import the latest SDNC certificate into ransim

   1. Download the SDNC cacerts from browser.

   2. Import the cacert file into the directory ran-simulator/ransim/docker/config/ransim/ and name it as 'jssecacerts'.(Sample step to import shown below)

      • keytool -import -noprompt -trustcacerts -alias SDNC -file <certsfile> -keystore /ransim/docker/config/ransim/jssecacerts -storepass changeit

3. Go to the ran-simulator/ransim directory and build the project : JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64/ mvn clean install -P docker
4. From docker folder, run the following after building the image locally and updating the SDNR IP address in the docker-compose file : docker-compose up

5. Run the following command to start the RAN network slice simulation

   curl -X POST http://localhost:8081/ransim/api/StartRanSliceSimulation -i

   To access the logs:
   docker exec –ti ransim bash
   policy@ransim:/tmp/ransim-install$ tail -f /opt/app/policy/servers/ransim/logs/ransim-rest.log 
6. Setup honeycomb (Refer HONEYCOMB section)
7. You can update/add/delete data(throughput, PRBs, maxNoConns etc) in mariadb container(credentials in docker-compose file) if required before generation of PM data.
8. Run the following command to generate Closed Loop PM data:
   1. curl -X POST http://localhost:8081/ransim/api/generateClosedLoopPmData -i
9. Run the following command to stop Closed Loop PM data(When enough PM data is generated): 
   1. curl -X GET http://localhost:8081/ransim/api/stopClosedLoopPmData -i
10. Access the DU honeycomb containers to view the logs:
    1. docker container logs -f 110
11. Check the folder(where SFTP setup is done) to view the generated PM data files.
12. Access VES, data file collector, dmaap-dr, PM mapper and slice-analysis logs to verify.

Honeycomb:

Repo URL: https://github.com/onap-oof-pci-poc/ran-sim

Pre-requisite: SDNC & Ransim controller should be up and running.

Build Steps:

If there is a new yang model, build the gnbsim-api with new yang model

Step 1: copy the new yang model to ~/ran-sim/hcsim-content/gnbsim/gnbsim-api/src/main/yang
Step 2: Go to directory ~/ran-sim/hcsim-content/gnbsim/gnbsim-api/
Step 3: Issue the build command "JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64/ mvn clean install -Dcheckstyle.skip"

For code changes in gnbsim-impl directory

Step 1: Do necessary code changes in gnbsim-impl
Step 2: Issue the build command "JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64/ mvn clean install -Dcheckstyle.skip"

Once the mvn build is done, build the docker image

Step 1: Go to directory ~/ran-sim/hcsim-content/gnbsim
Step 2: JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64/ mvn clean install -Dcheckstyle.skip

Step 3: Import the latest VES certificate into this directory gnbsim-distribution/ rename it as "jssecacerts"
Step 3: cp -r gnbsim-distribution/jssecacerts gnbsim-distribution/target/gnbsim-distribution-1.19.08-SNAPSHOT-hc/gnbsim-distribution-1.19.08-SNAPSHOT/
Step 4: cp -r gnbsim-distribution/honeycomb gnbsim-distribution/target/gnbsim-distribution-1.19.08-SNAPSHOT-hc/gnbsim-distribution-1.19.08-SNAPSHOT/
Step 5: cp -r gnbsim-distribution/Dockerfile gnbsim-distribution/target/gnbsim-distribution-1.19.08-SNAPSHOT-hc/gnbsim-distribution-1.19.08-SNAPSHOT/
Step 6: docker build -t gn gnbsim-distribution/target/gnbsim-distribution-1.19.08-SNAPSHOT-hc/gnbsim-distribution-1.19.08-SNAPSHOT/

Cross check the docker image
docker images | grep gn

To run the Netconf Servers:

1. cd ran-sim/hcsim-content/gnbsim/honeycomb
2. Update Ransim(ransimIp), honeycomb(hcIp) and VES IP address at hc/config/gnbsim.json
   1. {
        "ransimIp": "<IPAddress>",
        "ransimPort": 8081,
        "hcIp": "<IPAddress>",
        "hcPort": 2850,
        "endpoint": "ransim/RansimAgent",
        "vesEventListenerUrl": "https://<IPAddress>:30417/eventListener/v7"
      }
3. Update honeycomb  IP address at hc/docker-compose.yml file.
4. Step 2(Updating IP Address in config/gnbsim.json) and Step 3(Updating IP Address in docker-compose.yml file) is repeated for all folders (hc, hc-1, cucp1, cucp2, cuup1,cuup2, du1, du2, du3, du4, du5, du6)

      2. Run the up.sh script in ran-sim/hcsim-content/gnbsim/honeycomb. (wait for all the honeycomb servers to come up)

         After execution of the above script, these servers will be mounted in SDNC(check ransim and SDNR logs if the mount request is successful for all the servers). 

         Access the restconf page to view the mounted servers: https://<WorkerVMIpAddress>:30267/apidoc/explorer/index.html

Setup SFPT:

1. Run the below docker command in the VM where you have setup the netconf servers:

           docker run -v /home/$USER/upload:/home/$USER/sftptest/upload -p 2222:22 -d atmoz/sftp $USER:pass:1001

      2. Give chmod 777 to upload folder which will be created automatically in the home dir.

Smart Intent Guarantee based on Closed-loop (In Progress - will be updated soon)

ML-prediction-ms:

1. git clone "https://gerrit.onap.org/r/dcaegen2/services"
2. cd services/components/ml-prediction-ms
3. Update Ransim, Dmaap and ConfigDB IPAddress/Port details in services/components/ml-prediction-ms/ml-prediction-ms.config
4. Build docker image using: 'docker build -t ml-prediction-ms:latest .'
5. Run the ml-prediction-ms container.

RANSIM :

1. After starting the ml-prediction-ms container access the below logs to view the generated Intelligent slicing PM data.
2. To view the CUCP honeycomb containers  logs:
   1. docker container logs -f cucpserver1
3. Check the folder(where SFTP setup is done) to view the generated PM data files.
4. Access VES, data file collector, dmaap-dr, PM mapper, ml-prediction-ms and slice-analysis logs to verify.

Further References:

Issues faced during the RAN Simulator set up are documented in Setup related issues.