ONAP PODs failing in Casablanca Maintenance

 
1
0
-1

Hi Helen Chen


I tried installing ONAP R3 Maintenance with following flavors 

Openstack Pike

K8s Cluster of 13 VMs + 1 Rancher Node

flavors of VMs : 16 G RAM, 8 vCPUs & 70 G Disc


Currently status of pods are as follows:

Total pods: 250

Total Running/ Completed Pods : 217

Failed Number Of Pods : 33


See below screenshot of failed pods


dev-dmaap-dmaap-dr-prov-5c766b8d69-qrrrn                      0/1       CrashLoopBackOff   88         7h        

dev-contrib-netbox-app-provisioning-7gln5                     0/1       Error              0          7h        10.42.223.193   k8-node12   <none>

dev-portal-portal-db-config-7wgnd                             0/2       Error              0          7h        10.42.47.172    k8-node11   <none>

dev-portal-portal-db-config-8zhwq                             0/2       Error              0          7h        10.42.33.206    k8-node11   <none>

dev-portal-portal-db-config-hdttm                             0/2       Error              0          7h        10.42.65.42     k8-node5    <none>

dev-portal-portal-db-config-thts9                             0/2       Error              0          7h        10.42.65.112    k8-node11   <none>

dev-portal-portal-db-config-w5spq                             0/2       Error              0          7h        

10.42.224.233   k8-node9    <none>

dev-sdc-sdc-dcae-be-tools-65h9w                               0/1       Init:Error         0          6h        10.42.171.18    k8-node1    <none>

dev-sdc-sdc-dcae-be-tools-ffv6t                               0/1       Init:Error         0          7h        10.42.44.49     k8-node8    <none>

dev-sdc-sdc-dcae-be-tools-fssjw                               0/1       Init:Error         0          7h        


I tried debug system but not able to conclude exact reason.


Please help me on this.

Thanks

Abhay

    CommentAdd your comment...

    1 answer

    1.  
      1
      0
      -1

      Abhay Narayan Katare  Init:Error for sdc-dcae-be-tools is expected since the dependency job will not completed by that init timer expires. So, you can simply delete sdc-dcae-be-tools pods. 

      For portal try to see the following. There must a pods which is in completed state

      kubectl -n onap get pods -o wide | grep portal


      for dmaap-prov issue, you need to do following tweaks. This is for data-router only.

      apply the changes specified here manually https://gerrit.onap.org/r/#/c/79210/

      For more information look here  DMAAP-1065 - [DR] Casablanca - AAF certs expired on dmaap-dr-prov and dmaap-dr-node Closed

      and redeploy DMAAP.

      1. Abhay Narayan Katare

        Thanks a lot kranthi guttikonda. Let me try this and will get back to you for any further doubts

      2. Abhay Narayan Katare

        kranthi guttikonda we tried work around as suggested and still in the same state as above.


      3. kranthi guttikonda

        Abhay Narayan Katare Strange it worked for me. However latest comments include image1.0.9 and changes have been merged to casablanca branch. So, I would recommend try git clone the casablanca version and redeploy DMAAP. Make sure you remove all the pvc, pv and secrets for DMAAP before redeploying. 

      4. Abhay Narayan Katare

        kranthi guttikonda Thanks for your response , are you talking about Cloning from R3 Only rather than R3 Maintenance ? however as per my understanding R3 Maintenance is latest stable , Do changes has been cherry picked to R3 from maintenance ?. 

      5. kranthi guttikonda

        Abhay Narayan Katare clone R3 maintenance release. Make sure image: onap/dmaap/datarouter-node:1.0.9 in Values.yaml for dr-prov and dr-node charts.

      CommentAdd your comment...