- Submariner deployments can occasionally become tangled with dangling clusters or with connection issues.
- One way to resolve these is by deleting all of the pods in the
submariner-operator
namespace. - For Submariner on Azure, occasionally submariner may fall into a degraded state due to NAT being turned off. The fix for this is to patch the azure clusters to enable NAT.
- This can be done using the
fix_submariner_deployment.sh
script.
- define the clusters you wish to apply the fix to in the list
- The fix submariner deployment can perform a hard reset of all of the clusters provided.
- Replace the blank oc login command with either a token login, or kubeadmin credentials
- Uncomment the
# oc delete pods --all -n submariner-operator
line if you wish to refresh the pods. - Uncomment the
# oc -n submariner-operator scale deploy submariner-addon --replicas=0 # oc patch Submariner submariner --type=merge -p '{"spec":{"natEnabled": true}}' -n submariner-operator # oc -n submariner-operator scale deploy submariner-addon --replicas=1
lines should you wish to apply patches to azure clusters
- If a hub cluster fails and you need to gather credentials, or you wish to gather the kubeadmin credentials for all managed clusters you can use the
get_managedcluster_credentials.sh
script.
- To run the script simply define the credentials for the RHACM hub cluster you wish to gather credentials for.
- Run
get_managedcluster_credentials.sh