Rebooting your vRA environment

For me personally, it wasn’t really clear how you needed to reboot your vRA environment. Also when I followed the procedure I always had issues during startup.
So to clarify the process, I’ll write an article about it, that always helps myself to better understand it (or at least I hope so 🙂 )

This is the official VMware procedure => https://docs.vmware.com/en/vRealize-Automation/7.4/com.vmware.vra.prepare.use.doc/GUID-06E3645F-4360-4ACA-A721-620F9BE56DB1.html

Overview

This is the vRA environment I want to reboot.

VRA0010vRealize Automation Appliance (VIP)
VRA0011vRealize Automation Appliance (Node 1)
VRA0012vRealize Automation Appliance (Node 2)
VRA0013vRealize Automation Appliance (Node 3)
VRAW0010vRealize Automation IaaS Web (VIP)
VRAW0011vRealize Automation IaaS Web + Model Manager 1
VRAW0012vRealize Automation IaaS Web 2
VRAM0010vRealize Automation IaaS Manager (VIP)
VRAM0011vRealize Automation IaaS Manager + DEM Orchestrator 1
VRAM0012vRealize Automation IaaS Manager + DEM Orchestrator 2
CSQL0001MSSQL Database Server cluster name
RSQL0001MSSQL Database Server resource (client access point)
SQL0001vRealize Automation IaaS MSSQL Database (Node 1)
SQL0002vRealize Automation IaaS MSSQL Database (Node 2)
VRAD001vRealize Automation IaaS DEM Worker (Node1)
VRAD002vRealize Automation IaaS DEM Worker (Node2)
VRAP001vRealize Automation IaaS Proxy Agent (Node 1)
VRAP002vRealize Automation IaaS Proxy Agent (Node 2)
VRAP003vRealize Automation IaaS Proxy Agent (Node 3)
VRAP004vRealize Automation IaaS Proxy Agent (Node 4)

Shutdown vRA

Practically, these are the steps I followed translated towards my environment.

1. Shut down the Distributed Execution Manager Orchestrator and Workers and all vRealize Automation agents in any order and wait for all components to finish shutting down.

This confused me, certainly if you read in the next point that you need to shutdown VM’s that are running the Manager Service. The distributed Execution Manager Orchestrator holds the Manager service. So I’m assuming that they want us to shutdown the DEM orchestrator that doesn’t have the Manager Service running first.

How do you determine the primary / secondary IaaS Manager?
When you connect to the VAMI interface on one of your vRA appliances. Go to vRA settings => Cluster.
Scroll down till you see your IaaS Managers. The one where the ManagerService is running is the primary one.

Practically, what I did, was shutting down all my proxies & DEM workers. Wait untill the shutdown is finished and then shutdown the secondary IaaS Manager.

  • VRAP001 (IaaS Proxy Agents)
  • VRAP002 (IaaS Proxy Agents)
  • VRAP003 (IaaS Proxy Agents)
  • VRAP004 (IaaS Proxy Agents)
  • VRAD001 (IaaS DEM Worker)
  • VRAD002 (IaaS DEM Worker)
  • VRAM0012 (IaaS Manager + DEM Orchestrator)

2. Shut down virtual machines that are running the Manager Service and wait for the shutdown to finish.

Now we’re going to shutdown the primary IaaS Manager.

  • VRAM0011 (IaaS Manager + DEM Orchestrator)

3. Shut down all secondary Web nodes and wait for the shutdown to finish.

Because I have a distributed environment I first need to shutdown the secondary web node.

How do we determine the secondary or primary web node?
Again on the VAMI interface, go to vRA settings => Cluster.
Scroll down to where you find your web servers and search the one that holds the ModelManagerData. The server that holds this role is the primary web server.

  • VRAW0012 (IaaS Web)

4. Shut down the primary Web node, and wait for the shutdown to finish.

We already determined the primary web node in the previous point.

  • VRAW0011 (IaaS Web)

5. Shut down all secondary vRealize Automation appliance instances and wait for the shutdown to finish.

To determine the secondary vRA appliances I check the database settings on the VAMI interface. Go to vRA settings => Database.

The one that is flagged as master is the primary node so we’ll first reboot the other nodes first.

  • VRA0012 (vRealize Automation Appliance (Node 2))
  • VRA0013 (vRealize Automation Appliance (Node 3))

6. Shut down the primary vRealize Automation appliance and wait for the shutdown to finish.

The Master was determined in the previous point

  • VRA0011 (vRealize Automation Appliance (Node 2))

7. Shut down the MSSQL virtual machines in any order and wait for the shutdown to finish.

Because my SQL VM’s are in a Microsoft cluster I prefer to first shutdown the VM where the SQL isn’t running on. Once it’s powered off I shutdown the primary node that holds the SQL role.

Startup vRA

1. Start the MS SQL database machine.

Here we’re following the reverse order. First power on the primary node of the SQL cluster. Once it’s powered on you can now power on the secondary node.

2. In vSphere, start the master vRealize Automation appliance. Wait until the licensing service is running and REGISTERED in the master appliance management interface.

I can tell you that for vRA 7.4 this isn’t the way to go. This for the simple reason that your vRA simple will not start and you will see below errors on your primary appliance.

  • Your “Services” page will remain empty.
  • Your “vRA settings => Licensing” will give an error
  • Your “vRA settings => SSO” will give a status of 404

I have found that the best way is to start your primary vRA appliance and then after 1 min you start your secondary appliances.

  • VRA0011 (vRealize Automation Appliance (Node 1))
  • wait 1 min
  • VRA0012 (vRealize Automation Appliance (Node 2))
  • VRA0013 (vRealize Automation Appliance (Node 3))

3. Wait for the appliances to start, and verify that services are running and listed as REGISTERED in the appliance management interface.

You can check the status of your services on the VAMI interface, under the section services. It might take 15 or more minutes for appliances to start. The only service that will not start is the IAAS-service because he needs his windows machines powered-on for that.

4. Start the primary Web node and wait for the startup to finish.

Once you can logon to your windows machine you can continue to the next step.

  • VRAW0011 (IaaS Web)

5. Start all secondary Web nodes

  • VRAW0012 (IaaS Web)

Wait 5 minutes before you go to the next step

6. Start the primary Manager Service machine

  • VRAM0011 (IaaS Manager + DEM Orchestrator)

Wait 2 to 5 minutes.

7. Start the secondary Manager Service machine

  • VRAM0012 (IaaS Manager + DEM Orchestrator)

Wait 2 to 5 minutes.

8. Start the Distributed Execution Manager Orchestrator and Workers and all vRealize Automation proxy agents.

  • VRAP001 (IaaS Proxy Agents)
  • VRAP002 (IaaS Proxy Agents)
  • VRAP003 (IaaS Proxy Agents)
  • VRAP004 (IaaS Proxy Agents)
  • VRAD001 (IaaS DEM Worker)
  • VRAD002 (IaaS DEM Worker)

Wait for 2 to 5 minutes.
Now you’re ready to user your environment again.

Conclusion

I already executed – the steps provided – several times on a production environment so the procedure has been tested and validated several times 🙂