ONAP Beijing Release: Unable to access the Portal

 
1
0
-1

Hello Members, 

I am installing ONAP on Kubernetes on Rancher using the following links. 

https://onap.readthedocs.io/en/beijing/submodules/oom.git/docs/oom_setup_kubernetes_rancher.html#onap-on-kubernetes-with-rancher

https://onap.readthedocs.io/en/beijing/submodules/oom.git/docs/oom_quickstart_guide.html#quick-start-label

Here are some of the outputs of the commands entered on the Node

> netstat -nat | grep 30215
tcp6 0 0 :::30215 :::* LISTEN
> netstat -nat | grep 8989


> kubectl -n onap get services|grep "portal-app"

portal-app                 LoadBalancer   10.43.9.141     10.103.26.11                          8989:30215/TCP,8006:30213/TCP,8010:30214/TCP  


Contents of the etc/hosts file

10.103.26.11 portal.api.simpledemo.onap.org

10.103.26.11 vid.api.simpledemo.onap.org

10.103.26.11 sdc.api.fe.simpledemo.onap.org

10.103.26.11 portal-sdk.simpledemo.onap.org

10.103.26.11 policy.api.simpledemo.onap.org

10.103.26.11 aai.api.sparky.simpledemo.onap.org

10.103.26.11 cli.api.simpledemo.onap.org

10.103.26.11 msb.api.discovery.simpledemo.onap.org

The proxy settings on chrome have been disabled. 

When I try to access the portal using http://portal.api.simpledemo.onap.org:30215/ONAPPORTAL/applicationsHome the page is not loaded

also I had tried to <node-ip>:8989 and <node-ip>:30215 for each of the nodes but the apache tomcat home page is not getting displayed. 

I had the portal up a few weeks ago and a few days back it went down. I tried to use the suggestions here on these pages but to no avail. I also tried to redo the setup and still I face the issue. 

Also, I am observing since a few days that some of the nodes on the Rancher get disconnected after a certain period of time. 


Please note: I do not have floating IP's associated with the VMs. The VMs in Openstack are connected to the Internet via a router. 

Please, could you help me in sorting this out? 

Thanking you in anticipation. 





                              

    CommentAdd your comment...

    1 answer

    1.  
      1
      0
      -1

      Felix Dsouza, If it is some port access issues due to internal firewall specific rules, then we may not help much. However, please check the logs of the docker containers to see the Portal application has started correctly. Please find troubleshooting section in this wiki - Portal on OOM. This may help. 

      1. Felix Dsouza

        Hi Manoop, 

        The ingress ports are all open as well as the egress. The firewall for the system is all turned off. 

        The nodes are often getting disconnected and the IP for the kubectl -n onap get services|grep "portal-app" gets updated. I have allocated 4 VMs with each having RAM32GB, VCPUs16 VCPU and Disk200GB. 

        Below are the some of the logs

        >kubectl get services --namespace=onap | grep portal
        portal-app LoadBalancer 10.43.66.171 10.103.26.3 8989:30215/TCP,8006:30213/TCP,8010:30214/TCP 1d

        portal-cassandra ClusterIP 10.43.120.158 <none> 9160/TCP,7000/TCP,7001/TCP,7199/TCP,9042/TCP 1d

        portal-db ClusterIP 10.43.72.115 <none> 3306/TCP 1d

        portal-sdk NodePort 10.43.53.126 <none> 8080:30212/TCP 1d

        portal-widget ClusterIP 10.43.112.72 <none> 8082/TCP 1d

        portal-zookeeper ClusterIP 10.43.152.238 <none> 2181/TCP 1d

        sdnc-portal NodePort 10.43.91.232 <none> 8843:30201/TCP 1d


        > kubectl get pods --all-namespaces | grep portal
        onap dev-portal-app-54f8bc85d8-wcsqs 2/2 Running 8 1d
        onap dev-portal-cassandra-5ddbc59ffd-lnhwb 1/1 Unknown 0 18h
        onap dev-portal-cassandra-5ddbc59ffd-rtq5m 0/1 ContainerCreating 0 26m
        onap dev-portal-db-84f48ddccb-2gj9p 1/1 Running 1 14h
        onap dev-portal-sdk-756d9597b7-4gkt8 2/2 Running 2 1d
        onap dev-portal-widget-7dc7c978c7-pvw9t 0/1 PodInitializing 0 14h
        onap dev-portal-zookeeper-db466fc-5mrj6 1/1 Running 1 14h
        onap dev-sdnc-portal-64f5746fb6-nkpzm 0/1 Init:0/1 0 26m
        onap dev-sdnc-portal-64f5746fb6-q8nc5 0/1 Unknown 2 14h

        The portal was up a week ago when set up using the above tutorials. For a week there has been an issue.

      2. Manoop Talasila

        Hector Anapan-Lavalle/Gary Wu, do you see any reason why the 8989 or the nodeport is not accessible from the deployment?

      3. Gary Wu

        NodePorts should be available on all k8s hosts.  For the port 8989 you have to make sure you identify the right k8s host that the container is running on

      4. Felix Dsouza

        I did the reinstallation of ONAP. I have a question:  The one of the nodes as viewed on Rancher is in the Disconnected state. Is there a way to get the node active? This was the reason why I wasn't able to access the Portal. 

      CommentAdd your comment...