Nebula Operator manages NebulaGraph clusters on Kubernetes and automates tasks related to operating a NebulaGraph cluster. It evolved from NebulaGraph Cloud Service, makes NebulaGraph a truly cloud-native database.
See install/uninstall nebula operator .
$ kubectl create -f config/samples/nebulacluster.yamlA none ha-mode nebula cluster will be created.
$ kubectl get pods -l app.kubernetes.io/cluster=nebula
NAME READY STATUS RESTARTS AGE
nebula-graphd-0 1/1 Running 0 1m
nebula-metad-0 1/1 Running 0 1m
nebula-storaged-0 1/1 Running 0 1m
nebula-storaged-1 1/1 Running 0 1m
nebula-storaged-2 1/1 Running 0 1mSee client service for how to access nebula clusters created by the operator.
If you are working with kubeadm locally, create a nodePort
service and test that nebula is responding:
$ kubectl create -f config/samples/graphd-nodeport-service.yaml
$ nebula-console -u user -p password --address=192.168.8.26 --port=32236
2021/04/15 16:50:23 [INFO] connection pool is initialized successfully
Welcome to NebulaGraph!
(user@nebula) [(none)]>Destroy the nebula cluster:
$ kubectl delete -f config/samples/nebulacluster.yamlCreate a nebula cluster:
$ kubectl create -f config/samples/nebulacluster.yamlIn config/samples/nebulacluster.yaml the initial storaged replicas is 3.
Modify the file and change replicas from 3 to 5.
storaged:
resources:
requests:
cpu: "1"
memory: "1Gi"
limits:
cpu: "1"
memory: "1Gi"
replicas: 5
image: vesoft/nebula-storaged
version: v3.6.0
storageClaim:
resources:
requests:
storage: 2Gi
storageClassName: fast-disksApply the replicas change to the cluster CR:
$ kubectl apply -f config/samples/nebulacluster.yamlThe storaged cluster will scale to 5 members (5 pods):
$ kubectl get pods -l app.kubernetes.io/cluster=nebula
NAME READY STATUS RESTARTS AGE
nebula-graphd-0 1/1 Running 0 2m
nebula-metad-0 1/1 Running 0 2m
nebula-storaged-0 1/1 Running 0 2m
nebula-storaged-1 1/1 Running 0 2m
nebula-storaged-2 1/1 Running 0 2m
nebula-storaged-3 1/1 Running 0 5m
nebula-storaged-4 1/1 Running 0 5mSimilarly, we can decrease the size of the cluster from 5 back to 3 by changing the replicas field again and reapplying the change.
storaged:
resources:
requests:
cpu: "1"
memory: "1Gi"
limits:
cpu: "1"
memory: "1Gi"
replicas: 3
image: vesoft/nebula-storaged
version: v3.6.0
storageClaim:
resources:
requests:
storage: 2Gi
storageClassName: fast-disksWe should see that storaged cluster will eventually reduce to 3 pods:
$ kubectl get pods -l app.kubernetes.io/cluster=nebula
NAME READY STATUS RESTARTS AGE
nebula-graphd-0 1/1 Running 0 10m
nebula-metad-0 1/1 Running 0 10m
nebula-storaged-0 1/1 Running 0 10m
nebula-storaged-1 1/1 Running 0 10m
nebula-storaged-2 1/1 Running 0 10mIn addition, you can Install Nebula Cluster with helm.
Create a nebula cluster with the version specified (v3.6.0):
$ kubectl apply -f config/samples/nebulacluster.yaml
$ kubectl get pods -l app.kubernetes.io/cluster=nebula
NAME READY STATUS RESTARTS AGE
nebula-graphd-0 1/1 Running 0 25m
nebula-metad-0 1/1 Running 0 26m
nebula-storaged-0 1/1 Running 0 22m
nebula-storaged-1 1/1 Running 0 24m
nebula-storaged-2 1/1 Running 0 25mThe container image version should be v3.6.0:
$ kubectl get pods -l app.kubernetes.io/cluster=nebula -o jsonpath="{.items[*].spec.containers[*].image}" |tr -s '[[:space:]]' '\n' |sort |uniq -c
1 vesoft/nebula-graphd:v3.6.0
1 vesoft/nebula-metad:v3.6.0
3 vesoft/nebula-storaged:v3.6.0Now modify the file nebulacluster.yaml and change the version from v3.6.0 to v3.6.x:
Apply the version change to the cluster CR:
$ kubectl apply -f config/samples/nebulacluster.yamlWait few minutes. The container image version should be updated to v3.6.x:
$ kubectl get pods -l app.kubernetes.io/cluster=nebula -o jsonpath="{.items[*].spec.containers[*].image}" |tr -s '[[:space:]]' '\n' |sort |uniq -c
1 vesoft/nebula-graphd:v3.6.x
1 vesoft/nebula-metad:v3.6.x
3 vesoft/nebula-storaged:v3.6.xWarning:
Rolling upgrade must be within the same vX.Y version, if you need upgrade enterprise edition, please contact us.
See backup/restore nebula cluster .
If the minority of nebula components crash, the nebula operator will automatically recover the failure. Let's walk through this in the following steps.
Create a nebula cluster:
$ kubectl create -f config/samples/nebulacluster.yamlWait until pods are up. Simulate a member failure by deleting a storaged pod:
$ kubectl delete pod nebula-storaged-2 --nowThe nebula operator will recover the failure by creating a new pod nebula-storaged-2:
$ kubectl get pods -l app.kubernetes.io/cluster=nebula
NAME READY STATUS RESTARTS AGE
nebula-graphd-0 1/1 Running 0 15m
nebula-metad-0 1/1 Running 0 15m
nebula-storaged-0 1/1 Running 0 15m
nebula-storaged-1 1/1 Running 0 15m
nebula-storaged-2 1/1 Running 0 19sPlease refer to FAQ.md
Feel free to reach out if you have any questions. The maintainers of this project are reachable via:
Contributions are welcome and greatly appreciated.
nebula-operator refers to tidb-operator. They have made a very good product. We have a similar architecture, although the product pattern is different from the application scenario, we would like to express our gratitude here.
NebulaGraph is under the Apache 2.0 license. See the LICENSE file for details.