I have a DR cluster with 6 hosts (esxi 7.0.2, 17867351). Each host has 12 drive bays. I currently have a 4TB ssd (cache tier) and seven 6TB SAS hdds (capacity tier) in a disk group in each host, combined into a vsan datastore. We now need more space.
I was hoping to simply throw another 6TB sas drive in each host and expand the disk groups, but...disk groups apparently can't have more than 8 devices in them.
So I'm assuming I'll need to add an ssd and 3 hdds as another disk group in each host if I want to use the remaining drive bays (and don't want to go with an external drive array via iscsi or fc).
If this is the correct assumption, and I create a new disk group on each host, will I be able to then extend the current datastore with the 6 new disk groups, or will I have to create a new datastore with those 6 disk groups?
Also, if I use 3x12TB sas drives for a disk group, does the old "10% ssd cache size" still apply...so I'll need a 4TB ssd as the cache drive?
To expand the datastore of a vSAN it is possible to have diskgroups of different sizes and numbers of disks but it is not recommended because it could impact the performance of the vSAN infrastructure
For the sizing of the cache, the 10% of the space occupied by the VMs is valid, here are some links where they explain the question:
You can only have one vSAN Datastore per Cluster
For the expansion, I would also consider adding an ESXi node (with related disks of the same size) in which we have not only the space increase but also the performance increase
Or you can think of replacing the disks of all nodes with disks of greater capacity (this operation I have never done) I bring you this link: