vSphere 6.5 HA questions

[internal-admin]

Hello,

We are setting up vSphere HA with ESXi on 3 nodes.  What are the maximum number of HA VMs that can be run per host?  I looked at the Maximums Configuration and the number of 4 looks pretty low.  Does that mean I can only run 4 VMs per host? 

We are going to be running around 15 VMs distributed over 3 nodes.  If the second ESXi box goes down, can it temporarily start them up on the first and third host, while a technician is working on the problematic node?

For the hardware that's ordered, we can run ESXi 6.5U3 and are getting Essentials Plus.  What is the maximum vCPU per VM with Essentials Plus?  This is what was recommended to us up front, but no one mentioned maximums.  I'm worried we're going to hit a brick wall very soon if we do not get this planned/staged correctly.

Thank you.

[e_espinel]

Hello.

Within VMware, FT ensures availability by keeping VM copies on a separate host machine. With only HA configured, the hypervisor attempts to restart the VM on the same host cluster

FT is notable for the absence of downtime, while HA requires time to reboot. In addition, HA works on each cluster, while FT works on each virtual machine.

Please check the information in the following link

vSphere HA Checklist

https://docs.vmware.com/en/VMware-vSphere/6.5/com.vmware.vsphere.avail.doc/GUID-BA85FEC4-A37C-45BA-9...

Fault Tolerance Requirements, Limits, and Licensing

https://docs.vmware.com/en/VMware-vSphere/6.5/com.vmware.vsphere.avail.doc/GUID-57929CF0-DA9B-407A-B...

https://configmax.esp.vmware.com/guest?vmwareproduct=vSphere&release=vSphere%206.5%20Update%201&cate...

[internal-admin]

Hello and thanks for the reply.   It's alright for a host to boot from another ESXi host if the other node goes offline.  We have 3 Dell servers with access to an md3220 (via SAS cards/cables), and were told that HA is what we were looking for.  I'm kind of new to all of this and looking for some advice from a more senior level admin on this. 

I guess we're going with HA.  So would it be correct to assume that if a node goes down, HA checks for a heartbeat for a certain amount of time, then starts the VM on another node (of the 2 remaining ESXi nodes) right?  So we would configure host two and three to start n of the # of VM that failed, if node one goes offline?  This way, normally we have 3 running nodes and around 15 VMs total (5 per host), so if one node fails it would start them elsewhere.  Unless I've completely misunderstood HA to begin with?  This is what the people at VMWare had pointed us towards.

Thanks so much for your reply!  I'm trying to find the best solution in a small company, so we were advised to get VMWare Essentials Plus, so we could run HA in the configuration mentioned above. 

[depping · Accepted solution]

Yes, that is exactly how HA works. HA will restart the failed VMs as long as it has the resources to do so! Considering you only have 15 VMs in total, I don't expect this to be a problem, unless you have an extremely low amount of memory and CPU.

I wrote a book on HA if you want to get a better understanding, it covers all the basics, and all the deep inside stuff as well. you can find it here:L https://www.rubrik.com/resources/white-papers/19/clustering-deep-dive-ebook

the official documentation can be found here:

https://docs.vmware.com/en/VMware-vSphere/7.0/com.vmware.vsphere.avail.doc/GUID-63F459B7-8884-4818-8...