Can anyone provide more clarification on the differences between cache and storage tiers? The Microsoft documentation doesn't provide many details about how it works or what's the use cases for each. We would like all of our VMs to run on SSDs, however we
have also have a file server with lots of old data that is rarely accessed but still needs to be accessible on demand. Being able to put that data on HDDs instead of SSDs would be a significant cost savings.
We need about 1TB of data for VMs and about 6TB for file server, but most of that is old and will be access very infrequently. If we bought 2TB of SSD space and 6TB of HDD space, we could allocate 1TB of SSD only for VMs then a 2nd virtual disk with 1TB
of SSD and 6TB of HDD space. The VMs would always run on SSD, and the storage tiering service in S2D would automatically migrate the most used data on the 2nd virtual disk to the SSD, and move the least accessed files to the HDDs? Is that how tiering works?
Also, would I need the column number to match? If I had 4 SSDs and 4 HDDs, would I want to setup the virtual disk with 4 columns? Or 8?
Would tiering be a better option than cache in this situation? It seems like using cache would silly for our purposes because NVMe is not an option or a need for us right now, and caching SSDs with SSDs would slow down the storage. The documentation
shows that Journal SSDs will cache for other capacity SSDs, but I do not want that, I only want them to cache for HDDs. Is that possible?
https://docs.microsoft.com/en-us/windows-server/storage/storage-spaces/understand-the-cache#manual