Below is what I went through with installing vRA 7.6 in my test environment. I figured it would be good to document it in general, not that anybody has not done this before but it makes for a good basis for the setup in my environment.
The setup will be a minimal setup for now. Later there will be an enterprise setup also but I still need to deploy NSX in my environment to allow for the load balancers.
The setup will not be based on an absolute minimum. There will be 3 VMs initially:
VRA
SQL Server (2016) I will most likely write a bit about this separately and in a bit more detail.
and the IaaS host itself.
The Windows part will be based on Windows 2016.
On a longer basis the idea is to start with installing VRA in the management cluster for site 1, then deploy an agent in site 2 and slowly grow the setup a bit. Once NSX is setup I will move on to an enter price model. Below is a couple of screenshots showing the difference between the two setups:
For the the minimal installation what you need is really the three components mentioned above. For the Enterprise model it starts with the load balancer being an integral part and scales out from there.
Requirements wise for now I then need the three components. They need to be sized.
The appliance already comes ready sized so the two Windows boxes need some minimum requirements. If you are going with a single server to host SQL and IaaS then you really do need 4 CPUs and 8 GB RAM. In my case since I went with two different servers I put 2 CPUs for the IaaS, which arguably could be grown to 4 CPUs and 8 GB RAM. I went with 2 CPUs and 8 GB RAM for the SQL server. You can read more about the requirements in this pdf in chapter 8 (p. 19). Also you need a minimum of 40 GB harddisk space for the Windows guests.
Before starting the installation itself you need to deploy the VRA, which is done by downloading it from my.vmware.com and then deploying the ova to your vCenter.
The other thing I find quite useful is to take a copy of the Worksheet and add it in Excel. You can then gather all the information required for the installation there, but also gives you some documentation for later, if and when you need to troubleshoot the setup.
Here is a sample of such a file:
You can get a copy of the file from here: vRA_7_Worksheet_empty
A final note before starting on the installation:
I tend to take snapshots of all vms before starting the wizard, rather than later in process of the installation. I do this mainly because I then have 3 clean images to go back to. One draw-back is that I have to do everything I did between the snapshot and the point of failure. So far though I have rarely had to revert to the snapshots.
So assuming by now you have the required Windows box and a deployed VRA appliance, let us get started. One final warning, I have not put all the screenshots from the installation as some steps are requiring you to click next or simply enter a password.
From the IaaS host, log into the vra appliance (https://vra-appliance.domain.local:5480). the first time you do this you will get a wizard like the one below. I do suggest you follow the wizard as it does automate as much as possible.
Quite simple, press next to start the wizard.
Now you need to chose the size of the deployment
As mentioned earlier I am doing a minimal deployment at the moment so select that and make sure to check the box with “Install Infrastrcuture as a Service” also, then continue to the next screen where it is time to deploy the agent for the IaaS vm.
Download the “vCAC-IaaSManagementAgentSetup.msi” and run it on the IaaS server.
Note that if you are using a Windows Server 2019, you will need a different version of the client, which you can download from the VRA product page on my.vmware.com.
Go through the msi file, the first steps are just next next accept EULA and then you need to enter the information for the VRA appliance:
After the second screen above, click next and install.
After that return to the Wizard, where you should now see the IaaS server listed. By default the installation will use the ESXi time, I tend to change it to point to NTP servers.
What is important after adding the NTP servers is that you must change time. What is also important is that the “Time Offset” remains at 0. After pressing the Change Time settings you should see a message that states that the Time settings were updated on the top, as you can see here.
Now it is time to run the pre-requisites checker to see what is ok and what needs to fixed.
When the screen comes up press the “Run” button at the bottom of the screen. This may result in some reboots and eventually you should get a screen like above that states all pre-requisites are ok and that you can continue.
For the “vRealize Automation Host” screen you need to enter the automation host address. If you have DNS implemented you should be able to resolve it via the button automatically. Make sure you validate the settings whenever you can on the next pages. It means that the wizard checks that you have given data that are correct and saves you later from going back to a snapshot or worse, doing a re-installation.
For Single Sign-on you need to enter a password for the Single Sign-on and hit next.
Now it is time to enter the web host and also the service account that will be used for that site. Additionally you should also enter a password here for the SQL database to allow encryption.
After this it is time to configure the SQL server connection.
Note here that you need to add the port number if you change it from the default of 1433. The format is hostname.domain,local,port/instance.
Enter the database name and validate the settings. Hopefully you do not have too many issues here, else there is/will be a trouble shooting post soon also.
Next is the DEM settings, this should come up automatically and you can just validate and continue.
After this, the agent settings needs to be verified. The important part here is to remember the agent name. You need this when connecting to the vCenter. The default name is vCenter, in my case I changed it to someting more descriptive for me lab. Again remember to validate the settings.
For the vRealize certificates you need to either have a certificate ready to use or you can generate self-signed certificates, first for the VRA and after that for the web server.
Fill in Organisation, Org unit and country and hit the button to generate the certificates.
Then continue to the validation of the settings so far, hit the “Validation” button at the bottom of the screen and wait for a while. It can take up to 30 minutes depending on the size. For me it usually goes faster though, especially for minimal installations.
You should end up with a screen like this:
Now the Wizard recommends you to perform snapshots. As mentioned earlier I prefer to do them before starting this whole process.
Press next and you get the installation screen.
Note that the installation will take a while (up to 30 minutes). Start the installation by clicking on the “Install” button and let things happen. If all goes well you end up with everything being installed. If not then do not despair, there are some tips coming in a future post on common troubleshooting points that I have found from experience.
After some time you should end up with a screen like the one above here.
Click on next to continue to licensing, enter the key you have there and then continue to Telemetrics where you can join the Customer Experience Improvement Program (CEIP).
Finally you have the post installation options.
As it indicates, options are … well optional, I personally tend to skip them as I do not want content in my default tenant. I prefer to create that content myself after the installation. If you chose the configure the “initial content” you will be asked to supply a password for the user “configurationadmin” this user will not be created otherwise by the wizard.
In my case I go with “continue” to get the final screen:
That concludes this post, to come is the configuration of the default tenant and creating the first tenant.