Now that we have vCenter Server Appliance in-place and three fully operational nodes, let’s tie it all together in a neat little cluster.
This is a three step process:
- we need the create a datacenter
- add the nodes to the datacenter
- then create a cluster within that datacenter
- and lastly, move each node to the cluster
Go grab a cup of coffee (I know I will…) and let’s go:
1. Creating the datacenter
- Select the vCenter server and click on “Create Datacenter”
- In the next screen, enter a name for your datacenter and click OK
2. Adding nodes to the new datacenter
The next steps is to add all the nodes to the datacenter.
- Click “Add host”
- Enter either the IP address or the hostname of the node, Click “Next”
- Enter the local root account and password for the node, Click “Next”
- You’ll see a certificate warning, ensure you have the correct node, Click “Next”
- The dialog box will show a summary of the node to be added
- The next screen shows all the licenses that are available to the vCenter server. As I have already added the correct license to the node and vCenter, it shows up here. If you don’t have one, just select the “Evaluation License”, Click “Next”
- As I am using this as a test environment, I will not use lockdown. Sometimes I need to connect the oldskool fatclient or some other client., Click “Next”
- Binsfeldius is the only option here for me, I mean, the whole point is to put stuff in one place 🙂 , Click “Next”
- The last page shows a summary of all the stuff it found on the node. You see the networks that are configured and the connected storage options, Click “Finish”
Now repeat for the remaining nodes.
After adding the 2nd node, there will be error messages and warnings. This is fine and related to the settings for vSphere HA failover. These will disappear after the last node.
3. Creating the cluster
A new screen is now available and shows all the features and options within the datacenter.
Right click the new datacenter and create the cluster. There are a couple of initial settings to set here.
DRS or Distributed Resource Scheduler (DRS) is one of the main reasons to build a cluster like this. It basically will move VM’s around the nodes automatically and keep the cluster nice and balanced.
vSphere HA goes hand-in-hand with DRS and monitors the nodes for failures in which case it will move the VM’s to another node
- Put a tickmark in the DRS box and leave the slider in the middle for now.
- Put a tickmark in the vSphere HA box and in Enable Host monitoing
- Enable admission control
- Allow only failure for one host (due to the number of VM’s I will be running, there is only enough capacity for one node to fail)
- The other stuff we don’t need, yet.
4. Moving nodes to the cluster
Now that the cluster is successfully created we can add each of the nodes to the cluster.
- Right-click a node and select “move to”
- Select the freshly created cluster, click “OK”
- Select the first bullet and click “OK”
- Repeat for the remaining two nodes
That’s it. Now, there are a couple of minor issues to fix such as coredumps and management network redundancy.