With VMware Cloud Foundation (VCF) 9.0, Broadcom/VMware has really stepped up their game when it comes to storage capabilities. With VCF 9, VMware isn’t just shipping features — they’re executing on a long-term vision of unified, software-defined storage that spans clouds and moving towards a single storage platform (block, file, object).
There are several new features introduced in VSAN 9, however there is one feature that is stealing the spotlight, global deduplication in vSAN ESA. But that’s just the beginning. We’re also getting major improvements to vSAN File Services and for Kubernetes workloads. Let’s dive in.
First off, let’s go a bit more into vSAN ESA.
What is vSAN ESA?
VMware vSAN Express Storage Architecture also known as ESA is the next-generation storage architecture designed to take full advantage of high-performance NVMe storage, fast networking, and modern server hardware. Built with a new log-structured file system, vSAN ESA enables faster data ingestion and improved space efficiency. Compared to the original storage architecture (OSA), vSAN ESA delivers 2–5x better performance on the same hardware, offering scalable and highly efficient storage for modern workloads.
VMware also now has plans to bring these capabilities to AWS, Azure, and Google Cloud as well.
All of the new capabilities are now being built on top of vSAN ESA.
It is also however important to note that vVols are deprecated beginning with VVF/VCF 9.0 and will be fully disabled in a future VVF/VCF 9.x release.
Global Deduplication
This is arguably the biggest and most anticipated feature in the release. vSAN ESA now introduces global deduplication, which is a massive upgrade from the per-disk group dedupe we’ve had with vSAN OSA.
So what’s the big deal?
- Deduplication now happens across the entire cluster, not just within a disk group – meaning higher efficiency and better dedupe ratios.
- It’s post-process, so there’s no risk of write performance impact.
- Data layout is optimized so vSAN ESA can still read large, contiguous data blocks efficiently – even when deduped.
That said, it’s currently released as Limited Availability. You’ll need to request access, and there are some requirements (like no support for stretched clusters yet). Hopefully that will change soon. However VMware is stating that you can get up to 8x the space savings with this feature.
Updates to vSAN File Services
vSAN File Services has been around for a while, but with VCF 9.0 it finally gets the enterprise-grade treatment it deserves. You can now have up to 500 file shares per cluster – double the previous limit – and up to 100 SMB shares for Windows environments.
What makes this really powerful is the native integration:
- Fully managed via the vSphere Client.
- Supports both NFS v3/v4.1 and SMB with Kerberos authentication.
- Mount commands are shown directly in the UI – no guessing, no Googling.
- Load balancing and failover are handled automatically through containerized protocol services on each host.
- Supports hyperconverged, disaggregated, 2-node, and stretched clusters – with placement policies and site affinity.
File Services runs stateless containers on each host to handle the protocol layer, while the actual data is managed by vSAN’s Virtual Distributed File System. And yes – you get per-share performance metrics (IOPS, latency, throughput) and usage stats.
A few caveats: Snapshots are only available via API, and you can’t use NFS shares as VM datastores. Replication must be done with external tools like rsync or Robocopy.
Network Separation in vSAN Storage Clusters (Formerly vSAN Max)
This next feature may not be flashy, but it’s important: the ability to separate client and server traffic in vSAN Storage Clusters.
This means you can do things like:
- Route client traffic east-west on 100GbE within the rack
- Route server/storage traffic north-south over a 10GbE link
The benefits?
- Better performance
- Improved security
- Greater architectural flexibility
It’s a solid infrastructure improvement that’ll pay off in larger, high-throughput environments.
vSAN Data Protection + VMware Live Recovery
Backup and disaster recovery just got better too. With 9.0, you can now replicate VMs between clusters using vSAN ESA’s native snapshot technology – and we’re talking 200-deep snapshots with little to no performance impact.
Highlights:
- 1-minute RPO (!)
- No additional appliances – everything is now part of the VMware Live Recovery (VLR) appliance
- Uses the same UI and Protection Group concept introduced in 8.x
This makes replication and recovery a native part of the platform rather than something you bolt on afterward.
Stretched Cluster Enhancements
Two key improvements here:
- Site Maintenance Mode (SMM) – available via RPQ for OSA.
- Lets you put an entire site into maintenance mode with one click, instead of host-by-host.
- This is a huge operational win, especially in large environments (e.g. 10+10+1 setups).
- Manual Take Over (Limited Availability) – a new “last resort” tool if you’ve lost a site that was already in maintenance mode.
- Also designed to help recover data in scenarios where both a data site and the witness go down at the same time.
- It’s complex, so it’s only available through RPQ for now – and some training/understanding is needed before use.
Summary
This was just a look at some of the key storage-related features in VCF 9.0. There’s more under the hood as well in terms of new capabilities.
If you’re already using vSAN and thinking about moving to ESA, or if you’ve got hybrid or edge environments needing simplified file access – this release is absolutely worth checking out.
Stay tuned – and as always, test things in your lab environment before testing these new capabilities in production.
from StarWind Blog https://ift.tt/uk9ONda
via IFTTT
No comments:
Post a Comment