Permalink
46a8eaa Feb 21, 2018
1 contributor

Users who have contributed to this file

215 lines (163 sloc) 8.59 KB

Overview

This guide enables the integration, deployment, and management of GlusterFS containerized storage nodes in a Kubernetes cluster. This enables Kubernetes administrators to provide their users with reliable, shared storage.

Another important resource on this topic is the gluster-kubernetes project. It is focused on the deployment of GlusterFS within a Kubernetes cluster, and provides streamlined tools to accomplish this task. It includes a setup guide. It also includes a Hello World featuring an example web server pod using a dynamically-provisioned GlusterFS volume for storage. For those looking to test or learn more about this topic, follow the Quick-start instructions found in the main README for gluster-kubernetes

This guide is intended to demonstrate a minimal example of Heketi managing Gluster in a Kubernetes environment. It is meant to highlight the major components of this configuration and is not suitable for production.

Infrastructure Requirements

  • A running Kubernetes cluster with at least three Kubernetes worker nodes that each have at least one available raw block device attached (like an EBS Volume or a local disk).
  • The three Kubernetes nodes intended to run the GlusterFS Pods must have the appropriate ports opened for GlusterFS communication. Run the following commands on each of the nodes.
iptables -N HEKETI
iptables -A HEKETI -p tcp -m state --state NEW -m tcp --dport 24007 -j ACCEPT
iptables -A HEKETI -p tcp -m state --state NEW -m tcp --dport 24008 -j ACCEPT
iptables -A HEKETI -p tcp -m state --state NEW -m tcp --dport 2222 -j ACCEPT
iptables -A HEKETI -p tcp -m state --state NEW -m multiport --dports 49152:49251 -j ACCEPT
service iptables save

Client Setup

Heketi provides a CLI that provides users with a means to administer the deployment and configuration of GlusterFS in Kubernetes. Download and install the heketi-cli on your client machine.

Kubernetes Deployment

All the following files are located under extras/kubernetes.

  • Deploy the GlusterFS DaemonSet
$ kubectl create -f glusterfs-daemonset.json
  • Get node names by running:
$ kubectl get nodes
  • Deploy gluster container onto a specified node by setting the label storagenode=glusterfs on that node.
$ kubectl label node <...node...> storagenode=glusterfs

Repeat as needed. Verify that the pods are running on the nodes, at least three pods should be running.

$ kubectl get pods
  • Next we will create a service account for Heketi:
$ kubectl create -f heketi-service-account.json
  • We must now establish the ability for that service account to control the gluster pods. We do this by creating a cluster role binding for our newly created service account.
$ kubectl create clusterrolebinding heketi-gluster-admin --clusterrole=edit --serviceaccount=default:heketi-service-account
  • Now we need to create a Kubernetes secret that will hold the configuration of our Heketi instance. The configuration file must be set to use the kubernetes executor in order for the Heketi server to control the gluster pods. Beyond that, feel free to experiment with the configuration options.
$ kubectl create secret generic heketi-config-secret --from-file=./heketi.json
  • Next we need to deploy an initial Pod and a Service to access that pod. In the repo you cloned, there will be a heketi-bootstrap.json file.

Submit the file and verify everything is running properly as demonstrated below:

# kubectl create -f heketi-bootstrap.json
service "deploy-heketi" created
deployment "deploy-heketi" created

# kubectl get pods
NAME                                                      READY     STATUS    RESTARTS   AGE
deploy-heketi-1211581626-2jotm                            1/1       Running   0          35m
glusterfs-ip-172-20-0-217.ec2.internal-1217067810-4gsvx   1/1       Running   0          1h
glusterfs-ip-172-20-0-218.ec2.internal-2001140516-i9dw9   1/1       Running   0          1h
glusterfs-ip-172-20-0-219.ec2.internal-2785213222-q3hba   1/1       Running   0          1h
  • Now that the Bootstrap Heketi Service is running we are going to configure port-forwarding so that we can communicate with the service using the Heketi CLI. Using the name of the Heketi pod, run the command below:

kubectl port-forward deploy-heketi-1211581626-2jotm :8080

If local port 8080 is free on the system where you are running the commands, you can run the port-forward command so that it binds to 8080 for convenience:

kubectl port-forward deploy-heketi-1211581626-2jotm 8080:8080

Now verify that port forwarding is working by running a sample query against the Heketi Service. The command should have printed the local port that it will be forwarding from. Incorporate that into a URL to test the service, as demonstrated below:

curl http://localhost:57598/hello
Handling connection for 57598
Hello from Heketi

Lastly, set an environment variable for the Heketi CLI client so that it knows how to reach the Heketi Server.

export HEKETI_CLI_SERVER=http://localhost:57598

  • Next we are going to provide Heketi with information about the GlusterFS cluster it is to manage. We provide this information via a topology file. There is a sample topology file within the repo you cloned called topology-sample.json. Topologies specify what Kubernetes Nodes the GlusterFS containers are running on as well as the corresponding raw block device for each of the nodes.

    Make sure that hostnames/manage points to the exact name as shown under kubectl get nodes, and hostnames/storage is the ip address of the storage network.

    IMPORTANT: At this time, the topology file must be loaded using a version of heketi-cli that matches the server version. As a last resort the Heketi container comes with a copy of heketi-cli that can be accessed via kubectl exec ....

    Modify the topology file to reflect the choices you have made and then deploy it as demonstrated below:

heketi-client/bin/heketi-cli topology load --json=topology-sample.json
Handling connection for 57598
	Found node ip-172-20-0-217.ec2.internal on cluster e6c063ba398f8e9c88a6ed720dc07dd2
		Adding device /dev/xvdg ... OK
	Found node ip-172-20-0-218.ec2.internal on cluster e6c063ba398f8e9c88a6ed720dc07dd2
		Adding device /dev/xvdg ... OK
	Found node ip-172-20-0-219.ec2.internal on cluster e6c063ba398f8e9c88a6ed720dc07dd2
		Adding device /dev/xvdg ... OK
  • Next we are going to use Heketi to provision a volume for it to store its database:
# heketi-client/bin/heketi-cli setup-openshift-heketi-storage
# kubectl create -f heketi-storage.json

Pitfall: If heketi-cli reports an error of "No space" when running the setup-openshift-heketi-storage subcommand you may have inadvertently run topology load with mismatched versions of the server and heketi-cli. Stop the running Heketi pod (kubectl scale deployment deploy-heketi --replicas=0), manually remove any signatures from the storage block devices and then resume running a Heketi pod (kubectl scale deployment deploy-heketi --replicas=1). Then reload the topology with a matching version of heketi-cli and retry the step.

  • Wait until the job is complete then delete the bootstrap Heketi:
# kubectl delete all,service,jobs,deployment,secret --selector="deploy-heketi"
  • Create the long-term Heketi instance:
# kubectl create -f heketi-deployment.json
service "heketi" created
deployment "heketi" created
  • Now that this is done the Heketi db will persist in a GlusterFS volume and will not reset every time the Heketi pod is restarted.

    Use commands such as heketi-cli cluster list and heketi-cli volume list to confirm that the cluster established earlier exists and that Heketi is aware of the db storage volume created during the bootstrapping phase.

Usage Example

There are two ways to provision storage. The common way is to setup a StorageClass which lets Kubernetes automatically provision storage for a PersistentVolumeClaim submitted. Alternatively one can manually create and manage volumes (PVs) through Kubernetes or work directly with volumes from heketi-cli.

Please refer to the gluster-kubernetes hello world example for more information on storageClass configuration.