IBM Cloud Kubernetes: Container Storage Interface (CSI) plugin - Data collection guide

How to collect the CSI plugin logs:

1. At the primary host, create directory for the logs (you can change the directory name if you wish):

   $> mkdir logs  

2. Collect csi-controller logs

   $> kubectl logs -n kube-system ibm-block-csi-controller-0  --all-containers > logs/ibm-block-csi-controller
3. Collect csi-node logs

   $> nodepods=`kubectl get -n kube-system pod -l app.kubernetes.io/component=csi-node --output=jsonpath={.items..metadata.name}`
   $> for pod in $nodepods; do  kubectl logs -n kube-system --all-containers $pod > logs/$pod; done
4. Collect csi-operator logs

   $> operatorpod=`kubectl get pods --all-namespaces |grep ibm-block-csi-operator|awk '{print $2}'`
   $> kubectl logs $operatorpod -n kube-system > logs/operator
5. Describe detail from all csi components

   $> kubectl describe all -l csi -n kube-system > logs/describe_csi
6. If the PVC/s is not bounded, collect the PVC describe to see the events

   $> kubectl describe pvc [PVC-NAME] > logs/pvc_not_bounded
7. if Application is not Running, collect the Application describe to see the events

   $> kubectl describe pod [POD-NAME] > logs/pod_not_running
8. If the issue is related to IBM FlashSystem A9000, collect Xray, if the issue is related to IBM System Storage V7000, SVC or IBM FlashSystem FS9100, collect a SNAP.