All credit for this 'fork' to ibm-blockchain specifically IBM's Helm Chart ibm-blockchain-network
Minor tweaks to IBM's Helm Chart to deploy to Kubernetes Engine on Google Cloud Platform (GCP).
(Hyperledger) Fabric is the foundation of IBM Blockchain Service.
Medium Story explaining "Fabric on Google Cloud Platform"
- Create Kubernetes Engine cluster
- Create NFS Service (see below)
- Install Helm
- Install Tiller (see below)
- Clone and install Chart (see below)
The Chart shares configuration data in a volume across multiple Pods and requires read-write many. This (read-write many) is not possible with Google's foundational Persistent Disk (PD) but it is possible by running an NFS Service on Kubernetes atop PD:
Credit to Amine Jallouli for kubernetes-nfs-volume-on-gke:
If you'd prefer to use HDD rather than SSD, you need not apply this file to your cluster. If you would like to use SSD, you need to apply this file:
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: ssd
provisioner: kubernetes.io/gce-pd
parameters:
type: pd-ssd
#reclaimPolicy: Retain
Create and apply this file to a Kubernetes cluster to create a PersistentVolumeClaim, an NFS Deployment and a NFS Service. If you would prefer to use HDD rather than SSD change storageClassName here to be default rather than ssd. If you would prefer to use SSD, please ensure you've applied the previous file to your cluster before you apply this file:
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: nfs
spec:
storageClassName: ssd
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 2Gi
...
---
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
name: nfs
spec:
replicas: 1
selector:
matchLabels:
role: nfs
template:
metadata:
labels:
role: nfs
spec:
containers:
- name: nfs
image: gcr.io/google_containers/volume-nfs:0.8
ports:
- name: nfs
containerPort: 2049
- name: mountd
containerPort: 20048
- name: rpcbind
containerPort: 111
securityContext:
privileged: true
volumeMounts:
- mountPath: /exports
name: nfs
volumes:
- name: nfs
persistentVolumeClaim:
claimName: nfs
readOnly: false
...
---
apiVersion: v1
kind: Service
metadata:
name: nfs
spec:
ports:
- name: nfs
port: 2049
- name: mountd
port: 20048
- name: rpcbind
port: 111
selector:
role: nfs
...
If -- as is likely (encouraged) -- you're using an RBAC-based Kubernetes Engine cluster, this is a good way to install Helm's Tiller into the cluster after you've created both the Kubernets Engine cluster and installed Helm:
kubectl create serviceaccount tiller \
--namespace=kube-system
kubectl create clusterrolebinding tiller \
--clusterrole cluster-admin \
--serviceaccount=kube-system:tiller
helm init --service-account=tille
You're now ready to install the Helm chart.
Clone it if not already, dry-run it to ensure it looks correct and then install it:
git clone https://github.com/DazWilkin/ibm-blockchain-network.git
helm install --debug --dry-run ibm-blockchain-network
helm install ibm-blockchain-network
The last command will return something similar to:
NAME: funky-dragonfly
LAST DEPLOYED: Tue Jun 5 07:16:41 2018
NAMESPACE: default
STATUS: DEPLOYED
RESOURCES:
==> v1/Pod(related)
NAME READY STATUS RESTARTS AGE
funky-dragonfly-ibm-blockchain-network-ca-f68c596fc-zg2mw 0/1 ContainerCreating 0 0s
ibm-blockchain-network-debug-nfs-654fd6f4d4-ddwvt 0/1 ContainerCreating 0 0s
funky-dragonfly-ibm-blockchain-network-orderer-787984744b-hm8lv 0/1 ContainerCreating 0 0s
funky-dragonfly-ibm-blockchain-network-org1peer1-775fcd5d49j8k5 0/1 ContainerCreating 0 0s
funky-dragonfly-ibm-blockchain-network-org2peer1-98888c8b6kgpkc 0/1 ContainerCreating 0 0s
==> v1/PersistentVolume
NAME CAPACITY ACCESS MODES RECLAIM POLICY STATUS CLAIM STORAGECLASS REASON AGE
ibm-blockchain-shared-pvc 1Gi RWX Retain Bound default/ibm-blockchain-shared-pvc 0s
==> v1/PersistentVolumeClaim
NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
ibm-blockchain-shared-pvc Bound ibm-blockchain-shared-pvc 1Gi RWX 0s
==> v1/Service
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
ibm-blockchain-network-ca NodePort 10.39.248.78 <none> 7054:30000/TCP 0s
ibm-blockchain-network-orderer NodePort 10.39.242.214 <none> 31010:31010/TCP 0s
ibm-blockchain-network-org1peer1 NodePort 10.39.250.198 <none> 5010:30110/TCP,5011:30111/TCP 0s
ibm-blockchain-network-org2peer1 NodePort 10.39.242.138 <none> 5010:30210/TCP,5011:30211/TCP 0s
==> v1/Pod
NAME READY STATUS RESTARTS AGE
funky-dragonfly-ibm-blockchain-network-utils 0/3 ContainerCreating 0 0s
==> v1beta1/Deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
funky-dragonfly-ibm-blockchain-network-ca 1 1 1 0 0s
ibm-blockchain-network-debug-nfs 1 1 1 0 0s
funky-dragonfly-ibm-blockchain-network-orderer 1 1 1 0 0s
funky-dragonfly-ibm-blockchain-network-org1peer1 1 1 1 0 0s
funky-dragonfly-ibm-blockchain-network-org2peer1 1 1 1 0 0s
NOTES:
1. Get the application URL by running these commands:
export POD_NAME=$(kubectl get pods --namespace default -l "app=ibm-blockchain-network,release=funky-dragonfly" -o jsonpath="{.items[0].metadata.name}")
echo "Visit http://127.0.0.1:8080 to use your application"
kubectl port-forward $POD_NAME 8080:80
2. Next, create a channel and have your peers join by running these commands:
helm install stable/ibm-blockchain-channel
Enumerate deployed Charts:
helm list
Delete deployed Charts:
helm list will include the names of the charts that you may reference to delete charts:
helm delete [[NAME]]
You may delete the NFS Service in Kubernetes by delete-ing the files that you apply-ied earlier, perhaps (!):
kubectl delete --filename=nfs-deployment-service.yaml
kubectl delete --filename=ssd.yaml