Rumors on vSphere 5

The release of VMware vSphere 5.0 is almost visible on the horizon now.
In the online communities and when talking with colleagues the term ESXi 5 and/or vSphere 5 keep coming up.

The general expectation is that the release will be announced at VMworld 2011 in Las Vegas.
(no promise just a wild guess)

Version 4 of VMware vSphere/ESX(i) is quite suitable for most of our needs, but other players on the hypervisor market (Microsoft, Citrix) are closing the gap fast. So VMware needs to keep innovating to stay ahead and the reputation of VMware demands a new and improved version which will again stun the world. Packed with features with which Microsoft will ‘quick’ it’s customers again until they got it themselves in 2-3 years.

But fortunately the rumors on VMware vSphere 5 look very promising. Of course the configuration maximums will be improved again, not that anyone will come even close, but the added value must be in the new and improved features in my opinion.

So, what can we expect.

Configuration maximums:

• Support for 32 vCPU’s per VM (cur. 8);
• Support for 2TB+ LUN’s (cur. 2TB – 512b);
• Support for up to 1 TB of VM memory (cur. 255GB);
• Support for up to 512 VMs per host (cur. 320);
• Support for up to 2 TB or RAM per host (cur. 1TB RAM) .

New and improved features.

Main:

• Build on the vSphere ESXi hypervisor architecture
  This will be the first version of the VMware hypervisor which will only be available in a small footprint ESXi version. Of course this ESXi version will be available in two versions: embedded (flash, USB) and installable (disk(s)).
• Distributed Resource Scheduling (DRS) for Storage
  DRS for storage will enhance Storage vMotion in order to provide automatic load balancing for storage. Users will be able to define groups of data stores, called storage pods capable of automatic load balancing based on capacity, increasing storage utilization. Storage Distributed Resource Scheduler (DRS) will use Storage vMotion to perform automatic load balancing if a disk becomes overloaded. Storage DRS users will be able to define groups of data stores, called “storage pods,” that will automatically load-balance based on capacity. Users can then provision virtual machines (VMs) to specific storage pods rather than to specific data stores;
• vSphere Auto Deploy combining host profiles, Image Builder and PXE
  The feature will simplify the task of managing ESXi installation and upgrade for hundreds of machines. With this feature, new hosts can be automatically provisioned based on user-defined rules, and rebuilding a server to a clean state is as simple as a reboot;
• Host EUFI boot support
• Accelerator for specific use with View (VDI) workloads, providing a read cache optimized for recognizing, handling and deduplicating VDI client images.

Storage:

• Storage APIs
  Array Integration: Thin Provisioning enabling reclaiming blocks of a thin provisioned LUN on the array when a virtual disk is deleted;
• Improved version of the Cluster File System, VMFS5
  The new VMFS5 file system will improve scalability and performance. This is achieved by adding finer-grained SCSI locking controls;
• iSCSI user interface support
  New UI for managing iSCSI storage through the vSphere Client and support for Storage I/O control on Network File System storage volumes;
• Swap to SSD (Solid State Disk);
• Storage driven storage delivery based on the VMware-Aware Storage APIs;
• Storage vMotion snapshot support.

Network:

• Network I/O control for Virtual Machines
  This feature will reserve bandwidth for high-priority workloads in a cluster if that clusters network path is overloaded. This feature will also enable multi-tenant deployments, and will bridge physical and virtual QoS by complying with a new IEEE 802.1 VLAN tagging standard.;
• vNetwork Distributed Switch improvements providing improved visibility in VM traffic.

vCenter server:

• vCenter Server Appliance;
• vCenter Solutions Manager
  This providing a consistent interface to configure and monitor vCenter-integrated solutions developed by VMware and third parties.

Virtual machine:

• client-connected USB devices;
• USB 3.0 device support;
• Smart card reader support for VMs;
• UEFI virtual BIOS;
• Non-hardware accelerated 3D graphics for Windows Aero support;
• Apple Mac OS X Server 10.6 ‘Snow Leopard’ guest OS support.

High Availability:

• Improved VMware High Availability (HA) with Fault Domain Manager;
• All hosts in cluster can be primary nodes;
• Cluster also uses shared storage as a channel for heartbeat detection;
• Host-based replication for Site Recovery Manager
  This feature is designed for organizations that don’t use array-based storage replication, or use different types of storage at different sites. The replication will be asynchronous and be able to protect individual VMs.