Learn about the basics of Auto Deploy. Learn the basic function of Auto Deploy, and how it can be used to deploy an ESXi image to hosts.
- [Instructor] Our first lesson is a basic introduction to Auto Deploy. Now, Auto Deploy from VMware is a centralized server. Now, this is going to store the images of ESXi that are going to be used by hosts. This means that the vSphere software or ESXi image is never installed to a drive or static memory on the host at all, but is run directly from RAM, or random access memory. Now, Auto Deploy essentially holds our hypervisor, which is loaded by the host into RAM, where it can run directly.
No installation is needed. Traditional ESXi deployments require large amounts of administrative overhead in a large environment. Now, that greatly increases the provisioning time of a host. Now, this isn't too bad in small environments, right? We only have a few hosts, little amount of deployment time. But this may also be beneficial when deploying a customized host, or when we have a special configuration that needs to be done. But what happens when I take this to create a large data center of hosts that are fairly similar? For example, if I had to deploy 40 hosts in my environment using traditional means, I would have to install ESXi on 40 separate pieces of hardware to create a host and then I would have to configure those 40 hosts.
This is a fairly large time investment and could take days and days to complete. Now, on the other hand, if I used Auto Deploy, this same task could be accomplished much faster. By leveraging the centrally located image of ESXi, I could have all 40 hosts boot directly from the image I have stored on Auto Deploy server and this would save me hours of installation time. Now, I could also pair this with host profiles, which isn't a requirement for Auto Deploy at all, but is helpful.
I could then load configurations along with the ESXi hypervisor, cutting my time additionally. So, this is how it works. First, when the physical server is booting, this is our future host by the way, it contacts the DHCP server on the network for its information like in networking, or anything else that's been configured. Dynamic Host Configuration Protocol is most often used to provide an IP address to a node on the network. That's what we're familiar with, most of us, as administrators.
For Auto Deploy purposes, we will be configuring option 66 and option 67 of DHCP to point to a TFTP server. Now once the hardware has received the information from DHCP some of that information is going to point it to our TFTP server. This Trivial File Transfer Protocol server will contain a special file that we will store there that loads gPXE on our future host. Now, gPXE, or generic preboot execution environment, will allow us to load ESXi directly into memory.
Think of this as a preboot operating system. This will also point the hardware directly to an Auto Deploy server. The Auto Deploy server will then receive information from the hardware that will determine which image is delivered. This will often be the hardware model or even network information. That information will be checked against rules that are created with Auto Deploy to tie that information to a specific ESXi image that is stored on Auto Deploy.
Now, once the rules have been satisfied, the image tied to those rules is delivered to the hardware and now we have a host. ESXi is booted and, if set up properly, the Auto Deploy server will assign the host to vCenter Server.
- Deploying images with Auto Deploy GUI
- Creating a host profile
- Attaching hosts to a profile
- Modifying a host profile
- Detaching a host from a host profile
- Consolidating physical workloads using VMware Converter
- Installing vCenter Converter
- Converting a machine using vCenter Converter
- Troubleshooting conversion errors