In my last post, I talked about What’s new in vSphere 7 Update 1 when it comes to vSphere. This one covers the important bits (in my view) when it comes to vSAN. Due to the importance of vSAN for most vSphere environments, it’s good to see such a number of new features coming through. So, without further ado…
vSAN Direct Configuration
vSphere 7 Update 1 is introducing a new mechanism called the “Data Persistence Platform (DPp)” which will provide storage orchestration features, which sits between applications and the storage itself. While it requires application providers to create plug-ins to operate properly, it allows the platform to become extendable.
This change also enables vSAN to use “Direct Configuration” which allows directly connected storage devices connected to the hosts, to become available for consumption in the same way a vSAN is.
Doing so provides an excellent choice to admins who don’t need all the storage features that vSAN provides but still need some disk for its operations that can be accessed directly. As there’s a dependency on DPp I mentioned earlier, it will require the vSAN Ready Nodes from a provider that supports the new platform plug-in, which I am sure will start becoming available soon.
HCI Mesh
Now, what if you could “borrow” disk space when you needed it from vSAN neighbour? Sounds, useful? That’s exactly what the new feature “HCI Mesh” allows you to do”!
I am quite excited by this capability as if you think about it, not only it allows you to borrow capacity, it also effectively decouples the compute from storage and allows the flexible transfer of workloads.
vSAN Reserved Capacity
“Slack Space” is required for the smooth operation of vSAN and the recommended amount to protect is around 30%. For most organizations, it’s a lot of storage that they need to provision but can’t use. This even more painful for VMware Cloud on AWS deployments where an “on-demand” host is added automatically to the cluster to protect vSAN operations – something that does not sit well with customers.
That is why I am extremely excited about vSAN Reserved Capacity, which is the new name for that reservation, and it’s made up of two components: Host Rebuild Reserve + Operations Reserve.
In addition, VMware has made substantial changes to the way it operates and that will allow efficiencies in how that space is consumed. For that reason, comparatively less amount of space needs to be reserved, especially as the cluster size grows.
Naturally, some new configuration options have also been introduced into the GUI and it will generally expose more information to customers to help them better understand how the vSAN capacity is being used.
You may have noticed that Reserved Capacity is effectively switched off in the screenshot but that’s not a recommendation for typical use. There might be some use cases where switching off that reserve might make sense e.g. with service providers.
I am extremely glad about this addition as it will be a welcome change for everyone consuming vSAN as the primary storage.
Witness Host Consolidation
2-node vSAN was a major enhancement when it was made possible with the introduction of a witness host and many organisations were able to implement a cost-effective vSAN solution at branch offices.
Now that it has been around for a while, VMware is introducing a further dose of efficiency by allowing consolidation of the witness functionality for multiple vSAN clusters.
This is excellent news for larger organisations that have many sites being served by 2-node clusters but having to run many under-utilised witness appliances. This capability will probably be quite attractive to service providers too as it further reduces the cost of a smaller vSAN footprint.
Other Improvements
In addition to the above, there are a few other notable additions & improvements as well:
Compression Only: Now there’s a compression only option also available on vSAN. There are some workloads that don’t necessarily work well with deduplication. For such applications, a compression-only option is more suitable, and this option provides exactly that. Another side benefit of this option is that it reduces the failure domain as well.
Faster Host Restarts: Applicable to graceful restarts, this capability increases the efficiency of restarts by using a new “save and restore” workflow, which saves the in-memory metadata quickly, resulting in a shorter bring up time.
… and many more!
If there’s one platform update you have the time to delve deeper into, I would recommend picking vSAN for that read as there are a lot more updates that I could cover in this post. Most of them are such that every day vSAN users will be able to benefit from i.e. not just large organisations and all the operational improvements make this update a hugely significant one.
Leave A Comment