# Deploy Fluid on Your Kubernetes Cluster
# Prerequisites
- Git
- Kubernetes cluster（version >= 1.14）, and support CSI
- kubectl（version >= 1.14）
- Helm (opens new window)（version >= 3.0）
The following documents assume that you have installed all the above requirements.
For the installation and configuration of kubectl, please refer to here (opens new window).
For the installation and configuration of Helm 3, please refer to here (opens new window).
# How to Deploy
# Download Fluid Chart
You can download the latest Fluid installation package from Fluid Releases (opens new window).
Untar the Fluid package you just downloaded:
$ tar -zxf fluid.tgz
# Install Fluid with Helm
Create namespace:
$ kubectl create ns fluid-system
Install Fluid with:
$ helm install fluid fluid
NAME: fluid
LAST DEPLOYED: Fri Jul 24 16:10:18 2020
NAMESPACE: default
STATUS: deployed
REVISION: 1
TEST SUITE: None
The general format of the
helm installcommand is like:helm install <RELEASE_NAME> <SOURCE>. In the above command, the firstfluidmeans the release name, and the secondfluidspecified the path to the helm chart, i.e. the directory just unpacked.
# Upgrade Fluid to the latest version with Helm
If you have installed an older version of Fluid before, you can use Helm to upgrade it. Before upgrading, it is recommended to ensure that all components in the AlluxioRuntime resource object have been started completely, which is similar to the following state:
$ kubectl get pod
NAME READY STATUS RESTARTS AGE
hbase-fuse-chscz 1/1 Running 0 9h
hbase-fuse-fmhr5 1/1 Running 0 9h
hbase-master-0 2/2 Running 0 9h
hbase-worker-bdbjg 2/2 Running 0 9h
hbase-worker-rznd5 2/2 Running 0 9h
The command "helm upgrade" will not upgrade CRDs，we need to upgrade them manually：
$ tar zxvf fluid-0.5.0.tgz ./
$ kubectl apply -f fluid/crds/.
upgrade fluid：
$ helm upgrade fluid fluid/
Release "fluid" has been upgraded. Happy Helming!
NAME: fluid
LAST DEPLOYED: Fri Mar 12 09:22:32 2021
NAMESPACE: default
STATUS: deployed
REVISION: 2
TEST SUITE: None
The old version of controller will not terminate automatically, the new version will stay in the Pending status：
$ kubectl -n fluid-system get pods
NAME READY STATUS RESTARTS AGE
alluxioruntime-controller-56687869f6-g9l9n 0/1 Pending 0 96s
alluxioruntime-controller-5b64fdbbb-j9h6r 1/1 Running 0 3m55s
csi-nodeplugin-fluid-r6crn 2/2 Running 0 94s
csi-nodeplugin-fluid-wvhdn 2/2 Running 0 87s
dataset-controller-5b7848dbbb-rjkl9 1/1 Running 0 3m55s
dataset-controller-64bf45c497-w8ncb 0/1 Pending 0 96s
delete them manually：
$ kubectl -n fluid-system delete pod alluxioruntime-controller-5b64fdbbb-j9h6r
$ kubectl -n fluid-system delete pod dataset-controller-5b7848dbbb-rjkl9
We recommend you to update from v0.3 and v0.4. If you have an older version, you'd better to reinstall it.
# Check Status of Component
Check CRD used by Fluid:
$ kubectl get crd | grep data.fluid.io
alluxiodataloads.data.fluid.io 2020-07-24T06:54:50Z
alluxioruntimes.data.fluid.io 2020-07-24T06:54:50Z
datasets.data.fluid.io 2020-07-24T06:54:50Z
dataloads.data.fluid.io 2021-03-12T00:00:47Z
datasets.data.fluid.io 2021-03-12T00:00:47Z
jindoruntimes.data.fluid.io 2021-03-12T00:03:45Z
Check the status of pods:
$ kubectl get pod -n fluid-system
NAME READY STATUS RESTARTS AGE
alluxioruntime-controller-5dfb5c7966-mkgzb 1/1 Running 0 2d1h
csi-nodeplugin-fluid-64h69 2/2 Running 0 2d1h
csi-nodeplugin-fluid-tc7fx 2/2 Running 0 2d1h
dataset-controller-7c4bc68b96-26mcb 1/1 Running 0 2d1h
If the Pod status is as shown above, then Fluid is installed on your Kubernetes cluster successfully!
# Check version info of Component
When csi-nodeplugin, alluxioruntime-controller and dataset-controller start，they will print their own version info into logs.
If you installed with the charts provided by us，their version info will be fully consistent.
If you installed manually, their version info may be not consistent. You can view in turn:
$ kubectl logs csi-nodeplugin-fluid-tc7fx -c plugins -n fluid-system | head -n 9 | tail -n 6
$ kubectl logs alluxioruntime-controller-5dfb5c7966-mkgzb -n fluid-system | head -n 6
$ kubectl logs dataset-controller-7c4bc68b96-26mcb -n fluid-system | head -n 6
The printed logs are in the following format:
2020/10/27 10:16:02 BuildDate: 2020-10-26_14:04:22
2020/10/27 10:16:02 GitCommit: f2c3a3fa1335cb0384e565f17a4f3284a6507cef
2020/10/27 10:16:02 GitTreeState: dirty
2020/10/27 10:16:02 GoVersion: go1.14.2
2020/10/27 10:16:02 Compiler: gc
2020/10/27 10:16:02 Platform: linux/amd64
If the logs printed by Pod have been cleaned up, you can run the following command to view the version:
$ kubectl exec csi-nodeplugin-fluid-tc7fx -c plugins fluid-csi version -n fluid-system
$ kubectl exec alluxioruntime-controller-5dfb5c7966-mkgzb alluxioruntime-controller version -n fluid-system
$ kubectl exec dataset-controller-7c4bc68b96-26mcb dataset-controller version -n fluid-system
# Fluid use cases
For more use cases about Fluid, please refer to our demos:
- Speed Up Accessing Remote Files
- Cache Co-locality for Workload Scheduling
- Accelerate Machine Learning Training with Fluid
# Uninstall Fluid
For uninstalling fluid safely, we should check weather Custom Resource Objects about fluid have been deleted completely first:
kubectl get crds -o custom-columns=NAME:.metadata.name | grep data.fluid.io | sed ':t;N;s/\n/,/;b t' | xargs kubectl get --all-namespaces
If you confirm that all Custom resource objects about fluid have been deleted, you can safely uninstall fluid:
$ helm delete fluid
$ kubectl delete -f fluid/crds
$ kubectl delete ns fluid-system
The
fluidin commandhelm deletemeans the <RELEASE_NAME> during installation.
# Advanced Configuration
In some cloud vendors, the default mount root directory /runtime-mnt is not writable, so you have to modify the directory location
helm install fluid --set runtime.mountRoot=/var/lib/docker/runtime-mnt fluid