VMC High Availability

[9990374530]

Hi team,

As far as I know, Clusters can range from a minimum 3 hosts up to a maximum of 16 hosts per cluster.

> Bust what if we have 17 i3en hosts? How many cluster should I consider?

> I hope configuration for VMware part can still work if 2 hosts are not available at the same time?

> So if 2 hosts are dead the cluster would be able to support all the infrastructure?

> If we have 2 clusters, Does it works the same if each clusters looses 2 hosts?

> Do I need to use a policy that can survive 2 Failures when a cluster has six or more hosts? Or is it enabled by default?

vSphere High Availability ensures that if a host fails in an SDDC cluster, all virtual machines on the host are restarted on another host in the same cluster. 

Regards

PK

[ebernard · Accepted solution]

Hi PK,

Glad to see you again here!

Please find my answers below:

1) Clusters start at 2 nodes and max is 16 hosts

2) If you plan to deploy 17 nodes based on your requirements, I do recommend to split into 2 clusters (ie: 9 and 8 that will let you enough room to add hosts in the future if needed, in each cluster.

3) When you deploy your cluster, a storage policy (VSAN protection level, FTT)  is applied based on number of nodes, please see: https://docs.vmware.com/en/VMware-Cloud-on-AWS/services/com.vmware.vsphere.vmc-aws-manage-data-cente...  

4) If 2 hosts in a cluster in FTT=2, yes we will support and guarantee SLA if this applies to. Basically you look at the table and knowing which FTT you do want, you choose number of hosts.

5) If you split into many clusters (like for example 2), of course, SLA are at cluster level.  
if you want to change storage policy you can read this: https://docs.vmware.com/en/VMware-vSphere/7.0/com.vmware.vsphere.vsan.doc/GUID-F52F0AE9-FB31-4236-B5... 

6) vSphere HA is of course available in VMC and if an hosts fails we will replace it automatically and evacuate VMs from failed hosts. This is what we call "Host remediation". This is provided by default, with no additional fees.

Hope that answered to your questions and I can warmly recommend to contact your local VMware team to address all your questions in  the future as it will bring you an easier way to understand the service and all his mechanisms.

Cheers

[depping]

---

@Gianni654 wrote:

When the primary host can’t communicate with a secondary host over the management network, the primary uses datastore heartbeating to determine the state of that secondary.

---

Considering you only have vSAN, datastore heartbeating isn't actually enabled.